From aa31ab3e1282f9a5629168b54615349f110e5a78 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:48:07 -0700 Subject: [PATCH 01/18] contest: hw: write net.config for the performance tests The drivers/net/perf tests read the DUT and peer setup from net.config, like the drivers/net/hw ones. Write it for them too, so a hardware runner which builds the target can run them. Signed-off-by: Stanislav Fomichev --- contest/hw/hw_worker.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/contest/hw/hw_worker.py b/contest/hw/hw_worker.py index 55bb5784..b52f7a3d 100644 --- a/contest/hw/hw_worker.py +++ b/contest/hw/hw_worker.py @@ -487,7 +487,8 @@ def setup_test_interfaces(test_dir): if config_lines: config_content = '\n'.join(config_lines) + '\n' - for subdir in ['drivers/net', 'drivers/net/hw']: + for subdir in ['drivers/net', 'drivers/net/hw', + 'drivers/net/perf']: config_dir = os.path.join(test_dir, subdir) if os.path.isdir(config_dir): path = os.path.join(config_dir, 'net.config') From 3480dbf19117dd2dc93fcc00cc1b9232f3423e6d Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:48:07 -0700 Subject: [PATCH 02/18] contest: hw: run performance test cases one at a time Performance test cases are measurements. Run each case of a drivers/net/perf program in its own run_kselftest.sh process, with its own environment setup, so the cases cannot affect each other and each has clear boundaries. List the cases with -l, honoring the per-program arguments of run_kselftest.sh, which runs the environment setup of a program once more, and run each one with -t. Each case gets the whole per-test timeout, and a program which fails retries all of its cases. Signed-off-by: Stanislav Fomichev --- contest/hw/lib/runner.py | 121 +++++++++++++++++++++++++++-- contest/hw/tests/test_hw_worker.py | 110 ++++++++++++++++++++++++++ 2 files changed, 226 insertions(+), 5 deletions(-) diff --git a/contest/hw/lib/runner.py b/contest/hw/lib/runner.py index 8c748f3f..1236bfd9 100644 --- a/contest/hw/lib/runner.py +++ b/contest/hw/lib/runner.py @@ -5,6 +5,7 @@ import json import os import re +import shutil import signal import subprocess import time @@ -20,6 +21,7 @@ # How long to wait for a timed-out test to shut down gracefully after the # first (SIGINT) signal before we force-kill its whole process tree. GRACEFUL_KILL_WAIT = 60 +PERFORMANCE_TARGET = 'drivers/net/perf' # A test which dies with its *own* uncaught subprocess.TimeoutExpired never # runs its cleanup / __exit__, so netns, interfaces, XDP progs, qdiscs and @@ -270,7 +272,40 @@ def _kill_session(sid, sig): pass -def _run_one_test(test_dir, output_dir, target, prog, timeout): +def _test_args_name(prog): + """Return run_kselftest.sh's per-program argument variable name.""" + name = re.sub(r'[^A-Za-z0-9_]', '_', os.path.basename(prog)).upper() + return f'KSELFTEST_{name}_ARGS' + + +def _list_subtests(test_dir, target, prog, timeout): + """List Python kselftest cases without executing their test bodies.""" + test_path = os.path.join(test_dir, target, prog) + env = os.environ.copy() + env['NO_COLOR'] = '1' + # The arguments for run_kselftest.sh may filter the cases, split them the + # way it does + args = env.get(_test_args_name(prog), '').split() + completed = subprocess.run( + [test_path, '-l', *args], cwd=os.path.dirname(test_path), env=env, + capture_output=True, text=True, timeout=min(timeout, 60), check=False) + if completed.returncode: + error = completed.stderr.strip() or completed.stdout.strip() or \ + f'exit code {completed.returncode}' + raise RuntimeError(f'failed to list {target}:{prog} subtests: {error}') + cases = [] + for line in completed.stdout.splitlines(): + line = line.strip() + # Test setup may emit KTAP diagnostics before ksft_run() handles -l. + if line and not line.startswith('#'): + cases.append(line) + if not cases: + raise RuntimeError(f'{target}:{prog} did not list any subtests') + return cases + + +def _run_one_test(test_dir, output_dir, target, prog, timeout, + case_name=None): """Run a single test and save stdout/stderr. Returns (retcode, elapsed). On timeout the test is asked to shut down gracefully (SIGINT, which @@ -287,11 +322,17 @@ def _run_one_test(test_dir, output_dir, target, prog, timeout): is reached -- otherwise it is orphaned and lingers, poisoning the retry and later crashing on its own. """ + env = os.environ.copy() + if case_name is not None: + args_name = _test_args_name(prog) + env[args_name] = f'-t {case_name}' + t1 = time.monotonic() # pylint: disable=consider-using-with proc = subprocess.Popen( ['./run_kselftest.sh', '-t', f'{target}:{prog}'], cwd=test_dir, + env=env, stdout=subprocess.PIPE, stderr=subprocess.PIPE, start_new_session=True, @@ -343,6 +384,75 @@ def _run_one_test(test_dir, output_dir, target, prog, timeout): return retcode, elapsed +def _read_output(path, name): + with open(os.path.join(path, name), encoding='utf-8') as fp: + output = fp.read() + if output and not output.endswith('\n'): + output += '\n' + return output + + +def _run_performance_test(test_dir, output_dir, target, prog, timeout): + """Run each case in a performance test program in its own process.""" + try: + cases = _list_subtests(test_dir, target, prog, timeout) + except (OSError, RuntimeError, subprocess.TimeoutExpired) as error: + with open(os.path.join(output_dir, 'stdout'), 'w', + encoding='utf-8') as fp: + fp.write('') + with open(os.path.join(output_dir, 'stderr'), 'w', + encoding='utf-8') as fp: + fp.write(f'{error}\n') + return 1, 0.0 + + case_root = os.path.join(output_dir, '.subtests') + os.makedirs(case_root, exist_ok=True) + stdout = [] + stderr = [] + retcodes = [] + elapsed = 0.0 + + try: + for index, case_name in enumerate(cases): + print(f" {target}:{prog}: running subtest {case_name}") + case_dir = os.path.join(case_root, str(index)) + os.makedirs(case_dir) + retcode, case_elapsed = _run_one_test( + test_dir, case_dir, target, prog, timeout, + case_name=case_name) + retcodes.append(retcode) + elapsed += case_elapsed + stdout.append(_read_output(case_dir, 'stdout')) + stderr.append(_read_output(case_dir, 'stderr')) + + if _hit_timeout(case_dir): + break + finally: + shutil.rmtree(case_root, ignore_errors=True) + + with open(os.path.join(output_dir, 'stdout'), 'w', encoding='utf-8') as fp: + fp.write(''.join(stdout)) + with open(os.path.join(output_dir, 'stderr'), 'w', encoding='utf-8') as fp: + fp.write(''.join(stderr)) + + if any(retcode not in (0, 4) for retcode in retcodes): + retcode = next(retcode for retcode in retcodes + if retcode not in (0, 4)) + elif any(retcode == 0 for retcode in retcodes): + retcode = 0 + else: + retcode = 4 + return retcode, round(elapsed, 1) + + +def _execute_test(test_dir, output_dir, target, prog, timeout): + if target == PERFORMANCE_TARGET or \ + target.startswith(PERFORMANCE_TARGET + '/'): + return _run_performance_test( + test_dir, output_dir, target, prog, timeout) + return _run_one_test(test_dir, output_dir, target, prog, timeout) + + def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): """Execute kselftest in 'installed' form. @@ -350,7 +460,8 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): 1. Check dmesg for crash from previous test -- if found, stop 2. Check if test_name is in .attempted -- if so, skip (crash recovery) 3. Write test_name to .attempted + fsync (crash-safe bookkeeping) - 4. Run via: ./run_kselftest.sh -t : + 4. Run via: ./run_kselftest.sh -t :. Performance + programs run once per listed subtest. 5. Capture stdout/stderr, save to results_dir// 6. Drain dmesg output produced during the test, save to dmesg file 7. Save metadata (retcode, time, target, prog) to info file @@ -412,8 +523,8 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): os.makedirs(test_results_dir, exist_ok=True) # Run the test - retcode, elapsed = _run_one_test(test_dir, test_results_dir, - target, prog, timeout) + retcode, elapsed = _execute_test( + test_dir, test_results_dir, target, prog, timeout) with open(os.path.join(test_results_dir, 'stdout'), encoding='utf-8') as fp: stdout = fp.read() @@ -453,7 +564,7 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): if not skip_retry: retry_dir = os.path.join(results_dir, f'{dir_name}-retry') os.makedirs(retry_dir, exist_ok=True) - retry_retcode, _retry_elapsed = _run_one_test( + retry_retcode, _retry_elapsed = _execute_test( test_dir, retry_dir, target, prog, timeout) if _hit_timeout(retry_dir): tainted = True diff --git a/contest/hw/tests/test_hw_worker.py b/contest/hw/tests/test_hw_worker.py index 54fa36cb..ba8c7c9d 100644 --- a/contest/hw/tests/test_hw_worker.py +++ b/contest/hw/tests/test_hw_worker.py @@ -6,6 +6,7 @@ import subprocess import tempfile import unittest +from pathlib import Path from unittest import mock import sys @@ -18,6 +19,7 @@ _hit_timeout, _run_one_test, _known_bad_retry_decision) from lib.nipa import namify +from lib import runner # What a Python kselftest leaves in its output when one of *its* subprocess @@ -1108,5 +1110,113 @@ def test_unknown_driver_is_left_alone(self): self._run(None, (False, True)).assert_not_called() +class FakeProcess: + def __init__(self): + self.pid = 1234 + self.returncode = 0 + + def communicate(self, timeout=None): + return b'outer output\n', b'' + + +class TestPerformanceTestCases(unittest.TestCase): + def test_subtest_listing_ignores_ktap_diagnostics(self): + completed = mock.Mock( + returncode=0, + stdout=( + '# WARN: setup warning\n' + ' # Interface: eth0, driver: example\n' + 'case.one\n' + 'case.two\n' + ), + stderr='', + ) + with mock.patch.object(runner.subprocess, 'run', + return_value=completed): + cases = runner._list_subtests( + '/tests', runner.PERFORMANCE_TARGET, 'example.py', 10) + + self.assertEqual(cases, ['case.one', 'case.two']) + + def test_subtest_listing_honors_program_arguments(self): + completed = mock.Mock(returncode=0, stdout='case.two\n', stderr='') + with mock.patch.object(runner.subprocess, 'run', + return_value=completed) as run, \ + mock.patch.dict(runner.os.environ, { + 'KSELFTEST_EXAMPLE_PY_ARGS': '-t case.two'}): + cases = runner._list_subtests( + '/tests', runner.PERFORMANCE_TARGET, 'example.py', 10) + + self.assertEqual(cases, ['case.two']) + self.assertEqual(run.call_args.args[0], [ + os.path.join('/tests', runner.PERFORMANCE_TARGET, 'example.py'), + '-l', '-t', 'case.two']) + + def test_runs_one_subtest_through_program_arguments(self): + with tempfile.TemporaryDirectory() as temporary: + with mock.patch.object(runner.subprocess, 'Popen', + return_value=FakeProcess()) as popen: + retcode, _elapsed = runner._run_one_test( + temporary, temporary, runner.PERFORMANCE_TARGET, + 'example.py', 10, case_name='measured.case') + + self.assertEqual(retcode, 0) + environment = popen.call_args.kwargs['env'] + self.assertEqual(environment['KSELFTEST_EXAMPLE_PY_ARGS'], + '-t measured.case') + self.assertEqual( + Path(temporary, 'stdout').read_text(encoding='utf-8'), + 'outer output\n') + + def test_performance_program_runs_each_subtest_once(self): + with tempfile.TemporaryDirectory() as temporary: + test_dir = Path(temporary) / 'tests' + output_dir = Path(temporary) / 'output' + test_dir.mkdir() + output_dir.mkdir() + calls = [] + + def run_case(_test_dir, case_dir, _target, _prog, _timeout, + case_name=None): + calls.append(case_name) + Path(case_dir, 'stdout').write_text( + '# TAP version 13\n# 1..1\n' + f'# ok 1 {case_name}\n', encoding='utf-8') + Path(case_dir, 'stderr').write_text('', encoding='utf-8') + return 0, 1.0 + + with mock.patch.object(runner, '_list_subtests', + return_value=['case.one', 'case.two']), \ + mock.patch.object(runner, '_run_one_test', + side_effect=run_case): + retcode, elapsed = runner._run_performance_test( + str(test_dir), str(output_dir), + runner.PERFORMANCE_TARGET, 'example.py', 10) + + self.assertEqual(retcode, 0) + self.assertEqual(elapsed, 2.0) + self.assertEqual(calls, ['case.one', 'case.two']) + output = Path(output_dir, 'stdout').read_text(encoding='utf-8') + self.assertIn('# ok 1 case.one', output) + self.assertIn('# ok 1 case.two', output) + self.assertFalse(Path(output_dir, '.subtests').exists()) + + def test_performance_target_runs_cases_one_at_a_time(self): + with tempfile.TemporaryDirectory() as temporary: + with mock.patch.object(runner, '_run_performance_test', + return_value=(0, 1.0)) as performance, \ + mock.patch.object(runner, '_run_one_test', + return_value=(0, 1.0)) as regular: + runner._execute_test( + temporary, temporary, runner.PERFORMANCE_TARGET, + 'perf.py', 10) + runner._execute_test( + temporary, temporary, 'drivers/net/hw', + 'regular.py', 10) + + performance.assert_called_once() + regular.assert_called_once_with( + temporary, temporary, 'drivers/net/hw', 'regular.py', 10) + if __name__ == '__main__': unittest.main() From 8cc26077c0486eb2b12cf8726475c010538c9d2a Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:47:14 -0700 Subject: [PATCH 03/18] contest: remote: lib: import optional dependencies on first use Move namify() out of the fetcher, and import the remote helpers which need urllib3 or psutil on first use, so the result parsing helpers can be used without them, e.g. by tests or by tools running on a DUT. Do the same for Fetcher and NipaLifetime in contest/hw/lib/nipa.py, which re-exports the helpers for the hw code, so code running on a DUT can import it without the dependencies of the services. Signed-off-by: Stanislav Fomichev --- contest/hw/lib/nipa.py | 19 ++++++++++++++++--- contest/remote/lib/__init__.py | 33 ++++++++++++++++++++++++++++----- contest/remote/lib/fetcher.py | 10 ---------- contest/remote/lib/names.py | 11 +++++++++++ 4 files changed, 55 insertions(+), 18 deletions(-) create mode 100644 contest/remote/lib/names.py diff --git a/contest/hw/lib/nipa.py b/contest/hw/lib/nipa.py index ba273848..1f56531a 100644 --- a/contest/hw/lib/nipa.py +++ b/contest/hw/lib/nipa.py @@ -5,6 +5,9 @@ This module handles the sys.path setup so that other modules in contest/hw/ can simply ``from lib.nipa import ...`` without repeating the path manipulation. + +Fetcher and NipaLifetime need the dependencies of the services, they are +imported on first use, so the worker on a DUT can use the rest. """ import os @@ -23,6 +26,16 @@ from contest.remote.lib.results import result_from_indicators # noqa: E402, F401 from contest.remote.lib.results import parse_nested_tests # noqa: E402, F401 from contest.remote.lib.cbarg import CbArg # noqa: E402, F401 -from contest.remote.lib.fetcher import Fetcher # noqa: E402, F401 -from contest.remote.lib.fetcher import namify # noqa: E402, F401 -from core import NipaLifetime # noqa: E402, F401 +from contest.remote.lib.names import namify # noqa: E402, F401 + + +def __getattr__(name): + if name == 'Fetcher': + from contest.remote.lib.fetcher import Fetcher + globals()['Fetcher'] = Fetcher + elif name == 'NipaLifetime': + from core import NipaLifetime + globals()['NipaLifetime'] = NipaLifetime + else: + raise AttributeError(name) + return globals()[name] diff --git a/contest/remote/lib/__init__.py b/contest/remote/lib/__init__.py index 40a7ae4f..6e84643d 100644 --- a/contest/remote/lib/__init__.py +++ b/contest/remote/lib/__init__.py @@ -1,8 +1,31 @@ # SPDX-License-Identifier: GPL-2.0 -from .fetcher import Fetcher, namify -from .loadavg import wait_loadavg -from .vm import VM, new_vm +"""Remote runner helpers with lazy imports for optional dependencies.""" + from .cbarg import CbArg -from .crash import has_crash, extract_crash -from .results import guess_indicators, result_from_indicators, parse_nested_tests +from .crash import extract_crash, has_crash +from .names import namify +from .results import guess_indicators, parse_nested_tests +from .results import result_from_indicators + + +__all__ = [ + 'CbArg', 'Fetcher', 'VM', 'extract_crash', 'guess_indicators', 'has_crash', + 'namify', 'new_vm', 'parse_nested_tests', 'result_from_indicators', + 'wait_loadavg', +] + + +def __getattr__(name): + if name == 'Fetcher': + from .fetcher import Fetcher + globals()['Fetcher'] = Fetcher + elif name == 'wait_loadavg': + from .loadavg import wait_loadavg + globals()['wait_loadavg'] = wait_loadavg + elif name in ('VM', 'new_vm'): + from .vm import VM, new_vm + globals().update(VM=VM, new_vm=new_vm) + else: + raise AttributeError(name) + return globals()[name] diff --git a/contest/remote/lib/fetcher.py b/contest/remote/lib/fetcher.py index 06de3562..22f4296c 100644 --- a/contest/remote/lib/fetcher.py +++ b/contest/remote/lib/fetcher.py @@ -3,7 +3,6 @@ import datetime import json import os -import re import subprocess import time import urllib3 @@ -207,12 +206,3 @@ def _run_once(self): def run(self): while self.life.next_poll(): self._run_once() - - -def namify(what): - if not what: - return "no-name" - name = re.sub(r'[^0-9a-zA-Z]+', '-', what) - if name[-1] == '-': - name = name[:-1] - return name diff --git a/contest/remote/lib/names.py b/contest/remote/lib/names.py new file mode 100644 index 00000000..3e0ee84d --- /dev/null +++ b/contest/remote/lib/names.py @@ -0,0 +1,11 @@ +# SPDX-License-Identifier: GPL-2.0 + +"""Stable names for result URLs and database rows.""" + +import re + + +def namify(value): + if not value: + return 'no-name' + return re.sub(r'[^0-9a-zA-Z]+', '-', value).rstrip('-') From c39274e9e549d744f649182e92f256162750081d Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:20:36 -0700 Subject: [PATCH 04/18] contest: parse performance metrics from KTAP Parse JSON metric observations and policies from nested KTAP output. Attach them to the corresponding test case and keep initial and retry attempts separate. Ignore metrics which are not JSON objects with a name, and metrics with NaN or Infinity, which strict JSON consumers reject. Signed-off-by: Stanislav Fomichev --- AGENTS.md | 1 + contest/remote/lib/results.py | 35 +++++++++ contest/tests/__init__.py | 1 + contest/tests/fixtures/devmem-metrics.tap | 14 ++++ contest/tests/test_nested_metrics.py | 88 +++++++++++++++++++++++ 5 files changed, 139 insertions(+) create mode 100644 contest/tests/__init__.py create mode 100644 contest/tests/fixtures/devmem-metrics.tap create mode 100644 contest/tests/test_nested_metrics.py diff --git a/AGENTS.md b/AGENTS.md index 564a0a55..55042d2b 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -136,6 +136,7 @@ Run before committing changes to this repository's Python code. Unit tests (`unittest`-based, no `pytest` dependency required): ```console +$ python3 -m unittest discover -s contest/tests -p "test_*.py" $ python3 -m unittest discover -s contest/hw/tests -p "test_*.py" $ python3 -m unittest discover -s ml-agent/tests -p "test_*.py" ``` diff --git a/contest/remote/lib/results.py b/contest/remote/lib/results.py index 10dd7fc9..22d0e7db 100644 --- a/contest/remote/lib/results.py +++ b/contest/remote/lib/results.py @@ -6,9 +6,26 @@ hwksft (HW testing) to interpret test output. """ +import json import re +_METRIC_PREFIX = '# ktap-metric-json: ' + + +def _parse_metric_json(line): + try: + metric = json.loads(line[len(_METRIC_PREFIX):]) + # Strict JSON consumers, like the result database, reject NaN and + # Infinity + json.dumps(metric, allow_nan=False) + except (RecursionError, ValueError): + return None + if not isinstance(metric, dict) or not isinstance(metric.get('name'), str): + return None + return metric + + def guess_indicators(output): """Scan test output for pass/fail/skip indicator strings. @@ -60,6 +77,7 @@ def parse_nested_tests(full_run, namify_fn, prev_results=None): """ tests = [] nested_tests = False + pending_metrics = [] result_re = re.compile( r"(not )?ok (\d+)( \d+)?( -)? ([^#]*[^ ])( +# +)?([^ ].*)?$") @@ -72,13 +90,21 @@ def parse_nested_tests(full_run, namify_fn, prev_results=None): line = line[2:] else: nested_tests = False + pending_metrics = [] elif line.startswith("# TAP version "): nested_tests = True + pending_metrics = [] continue if not nested_tests: continue + if line.startswith(_METRIC_PREFIX): + metric = _parse_metric_json(line) + if metric is not None: + pending_metrics.append(metric) + continue + if line.startswith("ok "): result = "pass" elif line.startswith("not ok "): @@ -86,6 +112,9 @@ def parse_nested_tests(full_run, namify_fn, prev_results=None): else: continue + metrics = pending_metrics + pending_metrics = [] + match = result_re.match(line) if match is None: # A description is optional in TAP, while the stability names @@ -109,13 +138,19 @@ def parse_nested_tests(full_run, namify_fn, prev_results=None): for entry in prev_results: if entry['test'] == r['test']: entry['retry'] = result + if metrics: + entry['retry_metrics'] = metrics break else: # the first run didn't validate this test: add it to the list r['result'] = 'skip' r['retry'] = result + if metrics: + r['retry_metrics'] = metrics prev_results.append(r) else: + if metrics: + r['metrics'] = metrics tests.append(r) # return an empty list when there are prev results: no replacement needed diff --git a/contest/tests/__init__.py b/contest/tests/__init__.py new file mode 100644 index 00000000..f66554cd --- /dev/null +++ b/contest/tests/__init__.py @@ -0,0 +1 @@ +# SPDX-License-Identifier: GPL-2.0 diff --git a/contest/tests/fixtures/devmem-metrics.tap b/contest/tests/fixtures/devmem-metrics.tap new file mode 100644 index 00000000..4e775073 --- /dev/null +++ b/contest/tests/fixtures/devmem-metrics.tap @@ -0,0 +1,14 @@ +TAP version 13 +1..1 +# timeout set to 0 +# selftests: drivers/net/perf: devmem.py +# # WARN: net.ipv6.conf.eth0.keep_addr_on_down not set +# # WARN: LOCAL_V6 2001:db8::1 not configured with nodad +# # Interface: eth0, driver: mlx5_core +# TAP version 13 +# 1..1 +# # ktap-metric-json: {"goal":"higher","name":"throughput","value":389.127} +# ok 1 devmem.kperf +# # Totals: pass:1 fail:0 xfail:0 xpass:0 skip:0 error:0 +ok 1 selftests: drivers/net/perf: devmem.py +# Totals: pass:1 fail:0 xfail:0 xpass:0 skip:0 error:0 diff --git a/contest/tests/test_nested_metrics.py b/contest/tests/test_nested_metrics.py new file mode 100644 index 00000000..eee04b85 --- /dev/null +++ b/contest/tests/test_nested_metrics.py @@ -0,0 +1,88 @@ +# SPDX-License-Identifier: GPL-2.0 + +import json +import os +import unittest + +from contest.remote.lib.names import namify +from contest.remote.lib.results import parse_nested_tests + + +class TestNestedMetrics(unittest.TestCase): + def test_real_devmem_output(self): + """All metrics from the current devmem run belong to the kperf case.""" + fixture = os.path.join(os.path.dirname(__file__), 'fixtures', + 'devmem-metrics.tap') + with open(fixture, encoding='utf-8') as fp: + output = fp.read() + self.assertEqual(output.count('# # ktap-metric-json: '), 1) + for line in output.splitlines(): + marker = '# # ktap-metric-json: ' + if line.startswith(marker): + observation = json.loads(line[len(marker):]) + self.assertEqual(set(observation), {'name', 'value', 'goal'}) + + nested = parse_nested_tests(output, namify) + self.assertEqual(len(nested), 1) + self.assertEqual(nested[0]['test'], 'devmem-kperf') + self.assertEqual(nested[0]['result'], 'pass') + metrics = nested[0]['metrics'] + self.assertEqual(len(metrics), 1) + self.assertEqual(metrics[0]['name'], 'throughput') + self.assertAlmostEqual(metrics[0]['value'], 389.127) + self.assertEqual(metrics[0]['goal'], 'higher') + + + def test_retry_metrics_are_separate(self): + first = ('# TAP version 13\n' + '# # ktap-metric-json: {"name":"throughput",' + '"goal":"higher","value":300,"host":"local"}\n' + '# not ok 1 devmem.kperf\n') + retry = ('# TAP version 13\n' + '# # ktap-metric-json: {"name":"throughput",' + '"goal":"higher","value":320,"host":"local"}\n' + '# ok 1 devmem.kperf\n') + + nested = parse_nested_tests(first, namify) + parse_nested_tests(retry, namify, prev_results=nested) + + self.assertEqual(nested[0]['result'], 'fail') + self.assertEqual(nested[0]['retry'], 'pass') + self.assertEqual(nested[0]['metrics'][0]['value'], 300.0) + self.assertEqual(nested[0]['retry_metrics'][0]['value'], 320.0) + + def test_metric_fields_are_kept(self): + output = ( + '# TAP version 13\n' + '# # ktap-metric-json: {"name":"ring.rx_alloc_fail",' + '"goal":"zero","host":"remote","queue":3,"value":0}\n' + '# ok 1 devmem.kperf\n') + + nested = parse_nested_tests(output, namify) + + self.assertEqual(nested[0]['metrics'][0], { + 'name': 'ring.rx_alloc_fail', + 'goal': 'zero', + 'host': 'remote', + 'queue': 3, + 'value': 0, + }) + + def test_malformed_metrics_are_ignored(self): + output = ('# TAP version 13\n' + '# # ktap-metric-json: not-json\n' + '# # ktap-metric-json: []\n' + '# # ktap-metric-json: {"name":1}\n' + '# # ktap-metric-json: {"name":"x","value":NaN}\n' + '# # ktap-metric-json: {"name":"x","value":1e400}\n' + '# # ktap-metric-json: {"name":"x","value":' + + '[' * 100000 + ']' * 100000 + '}\n' + '# ok 1 devmem.kperf\n') + + nested = parse_nested_tests(output, namify) + + self.assertNotIn('metrics', nested[0]) + + +if __name__ == '__main__': + unittest.main() From 3c2be7ecb558cc9abbd50097ff2ec1492c164b0e Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:20:49 -0700 Subject: [PATCH 05/18] contest: summarize performance metrics per test case Build compact summaries and retain complete observations for each metric-bearing test-case attempt. Apply producer-selected transforms, aggregation, summaries and display metadata, and keep the summary value which regression checks compare. Label selftest metrics with the unit they report, or else with the one known for their name. Leave out values which are not finite numbers, skip attempts whose metrics cannot be summarized or stored as strict JSON, and store test cases sharing a sanitized name once. Signed-off-by: Stanislav Fomichev --- contest/backend/result_metrics.py | 255 +++++++++++++++++ contest/tests/test_nested_metrics.py | 13 + contest/tests/test_result_metrics.py | 407 +++++++++++++++++++++++++++ 3 files changed, 675 insertions(+) create mode 100644 contest/backend/result_metrics.py create mode 100644 contest/tests/test_result_metrics.py diff --git a/contest/backend/result_metrics.py b/contest/backend/result_metrics.py new file mode 100644 index 00000000..e24205ff --- /dev/null +++ b/contest/backend/result_metrics.py @@ -0,0 +1,255 @@ +# SPDX-License-Identifier: GPL-2.0 + +"""Helpers for extracting and summarizing metrics from a NIPA result.""" + +import json +import math + + +def _series_key(metric): + dimensions = { + key: value for key, value in metric.items() + if key not in ('time', 'value') + } + return json.dumps(dimensions, sort_keys=True, separators=(',', ':')) + + +def _nearest_rank(values, percentile): + rank = max(1, math.ceil(percentile * len(values))) + return values[rank - 1] + + +def _percentile(values, operation): + percentile = float(operation[1:]) + if not 0 <= percentile <= 100: + raise ValueError(f'percentile out of range: {operation}') + return _nearest_rank(values, percentile / 100) + + +def _summary_number(value): + return round(value, 6) if isinstance(value, float) else value + + +def _per_cpu_rate_samples(observations): + observations.sort(key=lambda observation: (observation[1]['time'], + observation[0])) + samples = [] + for (_prev_index, previous), (_index, current) in zip( + observations, observations[1:]): + elapsed = current['time'] - previous['time'] + if elapsed <= 0: + continue + rates = [ + (new - old) / elapsed + for old, new in zip(previous['value'], current['value']) + if old is not None and new is not None and new >= old + ] + if rates: + samples.append(rates) + return samples + + +def _scalar_rate_values(observations): + observations.sort(key=lambda observation: (observation[1]['time'], + observation[0])) + return [ + (current['value'] - previous['value']) / + (current['time'] - previous['time']) + for (_prev_index, previous), (_index, current) in zip( + observations, observations[1:]) + if current['time'] > previous['time'] and + current['value'] >= previous['value'] + ] + + +def _per_cpu_percent_totals(observations): + return [ + sum(value for value in metric['value'] if value is not None) / 100 + for _index, metric in observations + ] + + +# Units of the metrics the selftests report, for those which do not tell +_METRIC_UNITS = { + 'throughput': 'Gbps', +} + + +def _is_number(value): + # bool is an int, NaN and Infinity have no place in strict JSON + if isinstance(value, bool): + return False + if isinstance(value, float): + return math.isfinite(value) + return isinstance(value, int) + + +def _usable(metric, metric_type): + """Whether an observation has a value of its type to summarize.""" + value = metric.get('value') + if metric_type == 'per-cpu': + return isinstance(value, list) and \ + all(item is None or _is_number(item) for item in value) + return _is_number(value) + + +def summarize_metrics(metrics): + """Return one summary per metric series. + + Metrics of the selftests carry a name, a value, a goal and maybe a unit, + else NIPA may know theirs by name; scalars with their last value as the + summary, unless the metric says otherwise, as the system metrics of NIPA + do. Tests may print anything, values which are not finite numbers are + left out. + """ + series = {} + for index, metric in enumerate(metrics): + key = _series_key(metric) + if key not in series: + dimensions = { + name: value for name, value in metric.items() + if name not in ('time', 'value') + } + series[key] = { + 'dimensions': dimensions, + 'observations': [], + } + series[key]['observations'].append((index, metric)) + + summaries = [] + for entry in series.values(): + dimensions = entry['dimensions'] + metric_type = dimensions.get('type', 'scalar') + observations = [ + (index, metric) for index, metric in entry['observations'] + if _usable(metric, metric_type) + ] + if not isinstance(dimensions.get('name'), str) or not observations: + continue + base = dimensions.copy() + base['samples'] = len(observations) + + transform = base.get('transform') + aggregation = base.get('aggregation') + operation = base.setdefault('summarize', 'last') + # A unit the test reports wins over the one known for the name + unit = base.get('unit') + if not isinstance(unit, str): + unit = _METRIC_UNITS.get(base['name']) + if unit and 'display_label' not in base: + base['display_label'] = unit + + if metric_type == 'scalar': + if transform == 'rate': + values = _scalar_rate_values(observations) + else: + values = [metric['value'] + for _index, metric in observations] + elif metric_type == 'per-cpu': + if transform == 'rate': + rate_samples = _per_cpu_rate_samples(observations) + if aggregation == 'busy-core-equivalents': + values = [sum(sample) for sample in rate_samples] + else: + values = [value for sample in rate_samples + for value in sample] + elif aggregation == 'busy-core-equivalents': + values = _per_cpu_percent_totals(observations) + else: + values = [ + value + for _index, metric in observations + for value in metric['value'] + if value is not None + ] + base['cpus'] = max(len(metric['value']) + for _index, metric in observations) + else: + continue + + if not values: + continue + last_value = values[-1] + values.sort() + if operation and operation[0] == 'p': + base['summary'] = { + operation: _summary_number(_percentile(values, operation)), + } + elif operation == 'max': + base['summary'] = {'max': _summary_number(values[-1])} + elif operation == 'last': + base['summary'] = {'last': _summary_number(last_value)} + elif operation == 'total': + base['summary'] = {'total': _summary_number(sum(values))} + else: + continue + + if base.get('goal') in ('higher', 'lower', 'zero'): + base['regression_value'] = base['summary'][operation] + + base['value_count'] = len(values) + summaries.append(base) + + return summaries + + +def extract_result_metrics(data): + """Return one summarized/full metric record per test-case attempt.""" + rows = [] + # Distinct test cases may share a sanitized name, keep the first one + stored = set() + + for test in data.get('results') or []: + group = test.get('group') + test_name = test.get('test') + if not group or not test_name: + continue + + for case in test.get('results') or []: + subtest = case.get('test') + if not subtest: + continue + + base = { + 'group': group, + 'test': test_name, + 'subtest': subtest, + 'link': test.get('link'), + } + attempts = ( + (0, case.get('result'), case.get('metrics', [])), + (1, case.get('retry'), case.get('retry_metrics', [])), + ) + for attempt, result, metrics in attempts: + if not isinstance(result, str) or not isinstance(metrics, list): + continue + result = result.lower() + if result not in ('pass', 'fail', 'skip'): + continue + + key = (group, test_name, subtest, attempt) + if not metrics or key in stored: + continue + # Metrics come from the tests, which may print anything + try: + summary = summarize_metrics(metrics) + # Strict JSON consumers, like the result database, + # reject NaN and Infinity + json.dumps([summary, metrics], allow_nan=False) + except (AttributeError, LookupError, TypeError, ValueError, + ArithmeticError): + continue + if not summary: + continue + + stored.add(key) + row = base.copy() + row.update({ + 'short_json': summary, + 'full_json': metrics, + 'attempt': attempt, + 'result': result, + }) + rows.append(row) + + return rows diff --git a/contest/tests/test_nested_metrics.py b/contest/tests/test_nested_metrics.py index eee04b85..523dbd2b 100644 --- a/contest/tests/test_nested_metrics.py +++ b/contest/tests/test_nested_metrics.py @@ -4,6 +4,7 @@ import os import unittest +from contest.backend.result_metrics import extract_result_metrics from contest.remote.lib.names import namify from contest.remote.lib.results import parse_nested_tests @@ -32,6 +33,18 @@ def test_real_devmem_output(self): self.assertAlmostEqual(metrics[0]['value'], 389.127) self.assertEqual(metrics[0]['goal'], 'higher') + rows = extract_result_metrics({'results': [{ + 'group': 'selftests-drivers-net-perf', + 'test': 'devmem-py', + 'result': 'pass', + 'link': ('https://netdev.bots.linux.dev/results/123456/' + 'test-outputs/0-devmem-py/'), + 'results': nested, + }]}) + self.assertEqual(len(rows), 1) + self.assertEqual(rows[0]['subtest'], 'devmem-kperf') + self.assertEqual(len(rows[0]['short_json']), 1) + self.assertEqual(len(rows[0]['full_json']), 1) def test_retry_metrics_are_separate(self): first = ('# TAP version 13\n' diff --git a/contest/tests/test_result_metrics.py b/contest/tests/test_result_metrics.py new file mode 100644 index 00000000..064fd639 --- /dev/null +++ b/contest/tests/test_result_metrics.py @@ -0,0 +1,407 @@ +# SPDX-License-Identifier: GPL-2.0 + +import unittest + +from contest.backend.result_metrics import extract_result_metrics + + +class TestResultMetrics(unittest.TestCase): + def test_extracts_initial_and_retry_rows(self): + data = { + 'results': [{ + 'group': 'selftests-drivers-net-hw', + 'test': 'devmem-py', + 'link': ('https://netdev.bots.linux.dev/results/123456/' + 'test-outputs/0-devmem-py/'), + 'results': [{ + 'test': 'devmem-kperf', + 'result': 'fail', + 'retry': 'pass', + 'metrics': [{ + 'name': 'throughput', 'goal': 'higher', + 'value': 300.0, 'host': 'local', + }], + 'retry_metrics': [{ + 'name': 'throughput', 'goal': 'higher', + 'value': 320.0, 'host': 'local', + }], + }], + }], + } + + rows = extract_result_metrics(data) + + self.assertEqual(len(rows), 2) + self.assertEqual(rows[0], { + 'group': 'selftests-drivers-net-hw', + 'test': 'devmem-py', + 'subtest': 'devmem-kperf', + 'link': ('https://netdev.bots.linux.dev/results/123456/' + 'test-outputs/0-devmem-py/'), + 'short_json': [{ + 'name': 'throughput', + 'goal': 'higher', + 'host': 'local', + 'display_label': 'Gbps', + 'summarize': 'last', + 'samples': 1, + 'summary': {'last': 300.0}, + 'regression_value': 300.0, + 'value_count': 1, + }], + 'full_json': [{ + 'name': 'throughput', + 'goal': 'higher', + 'host': 'local', + 'value': 300.0, + }], + 'attempt': 0, + 'result': 'fail', + }) + self.assertEqual(rows[1]['short_json'][0]['summary']['last'], 320.0) + self.assertEqual(rows[1]['full_json'][0]['value'], 320.0) + self.assertEqual(rows[1]['attempt'], 1) + self.assertEqual(rows[1]['result'], 'pass') + + def test_short_json_summarizes_every_series(self): + metrics = [ + {'name': 'throughput', 'type': 'scalar', 'host': 'local', + 'time': 2.0, 'value': 320.0, 'summarize': 'p50'}, + {'name': 'throughput', 'type': 'scalar', 'host': 'local', + 'time': 1.0, 'value': 300.0, 'summarize': 'p50'}, + {'name': 'cpu.utilization', 'type': 'per-cpu', 'host': 'local', + 'time': 1.0, 'value': [10.0, None, 90.0], + 'aggregation': 'busy-core-equivalents', 'summarize': 'p90'}, + {'name': 'cpu.utilization', 'type': 'per-cpu', 'host': 'local', + 'time': 2.0, 'value': [100.0, 20.0, 30.0], + 'aggregation': 'busy-core-equivalents', 'summarize': 'p90'}, + ] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + row = extract_result_metrics(data)[0] + + self.assertEqual(row['full_json'], metrics) + self.assertEqual(row['short_json'], [ + { + 'name': 'throughput', 'type': 'scalar', 'host': 'local', + 'samples': 2, 'value_count': 2, 'summarize': 'p50', + 'display_label': 'Gbps', 'summary': {'p50': 300.0}, + }, + { + 'name': 'cpu.utilization', 'type': 'per-cpu', + 'host': 'local', 'samples': 2, 'value_count': 2, + 'aggregation': 'busy-core-equivalents', + 'summarize': 'p90', 'summary': {'p90': 1.5}, + 'cpus': 3, + }, + ]) + + def test_metric_name_does_not_imply_an_aggregation_policy(self): + metrics = [ + {'name': 'cpu.time.unconfigured', 'type': 'per-cpu', + 'host': 'local', 'time': 1.0, 'value': [0.0, 1.0], + 'summarize': 'p50'}, + {'name': 'cpu.time.unconfigured', 'type': 'per-cpu', + 'host': 'local', 'time': 2.0, 'value': [1.0, 3.0], + 'summarize': 'p50'}, + ] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + summary = extract_result_metrics(data)[0]['short_json'][0] + + self.assertEqual(summary, { + 'name': 'cpu.time.unconfigured', 'type': 'per-cpu', + 'host': 'local', 'samples': 2, 'cpus': 2, + 'summarize': 'p50', 'summary': {'p50': 1.0}, + 'value_count': 4, + }) + + def test_producer_metadata_drives_rate_and_cpu_aggregation(self): + metrics = [ + {'name': 'custom.cpu.counter', 'type': 'per-cpu', + 'host': 'local', 'time': 1.0, 'value': [0.0, 1.0, None], + 'transform': 'rate', + 'aggregation': 'busy-core-equivalents', 'summarize': 'p75', + 'display_range': {'min': 0, 'max': 100}, + 'display_scale': 100, + 'display_label': 'Percent of one CPU'}, + {'name': 'custom.cpu.counter', 'type': 'per-cpu', + 'host': 'local', 'time': 3.0, 'value': [1.0, 2.5, 3.0], + 'transform': 'rate', + 'aggregation': 'busy-core-equivalents', 'summarize': 'p75', + 'display_range': {'min': 0, 'max': 100}, + 'display_scale': 100, + 'display_label': 'Percent of one CPU'}, + ] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + summary = extract_result_metrics(data)[0]['short_json'][0] + + self.assertEqual(summary, { + 'name': 'custom.cpu.counter', 'type': 'per-cpu', 'host': 'local', + 'samples': 2, 'transform': 'rate', + 'aggregation': 'busy-core-equivalents', 'summarize': 'p75', + 'display_range': {'min': 0, 'max': 100}, + 'display_scale': 100, + 'display_label': 'Percent of one CPU', + 'cpus': 3, 'summary': {'p75': 1.25}, + 'value_count': 1, + }) + + def test_percentile_metadata_summarizes_one_sample(self): + metrics = [{ + 'name': 'custom.throughput', 'type': 'scalar', 'host': 'local', + 'time': 1.0, 'value': 42, 'summarize': 'p75', + }] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + summary = extract_result_metrics(data)[0]['short_json'][0] + + self.assertEqual(summary, { + 'name': 'custom.throughput', 'type': 'scalar', 'host': 'local', + 'samples': 1, 'summarize': 'p75', + 'summary': {'p75': 42}, 'value_count': 1, + }) + + def test_selftest_metrics_summarize_their_value(self): + metrics = [ + {'name': 'rr.latency.p99', 'goal': 'lower', 'host': 'remote', + 'unit': 'us', 'value': 25}, + {'name': 'ring.rx_alloc_fail', 'goal': 'zero', 'host': 'remote', + 'value': 0}, + ] + + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + self.assertEqual(extract_result_metrics(data)[0]['short_json'], [ + { + 'name': 'rr.latency.p99', 'goal': 'lower', 'host': 'remote', + 'unit': 'us', 'display_label': 'us', 'samples': 1, + 'summarize': 'last', + 'summary': {'last': 25}, 'regression_value': 25, + 'value_count': 1, + }, + { + 'name': 'ring.rx_alloc_fail', 'goal': 'zero', + 'host': 'remote', 'samples': 1, 'summarize': 'last', + 'summary': {'last': 0}, 'regression_value': 0, + 'value_count': 1, + }, + ]) + + def test_reported_unit_wins_over_known_one(self): + metrics = [{'name': 'throughput', 'goal': 'higher', 'unit': 'Mbps', + 'value': 940}] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + summary = extract_result_metrics(data)[0]['short_json'][0] + + self.assertEqual(summary['unit'], 'Mbps') + self.assertEqual(summary['display_label'], 'Mbps') + + def test_scalar_counter_metadata_drives_rate_percentile(self): + metrics = [ + {'name': 'custom.counter', 'type': 'scalar', 'host': 'local', + 'time': 0.0, 'value': 0, 'transform': 'rate', + 'summarize': 'p50'}, + {'name': 'custom.counter', 'type': 'scalar', 'host': 'local', + 'time': 1.0, 'value': 2, 'transform': 'rate', + 'summarize': 'p50'}, + {'name': 'custom.counter', 'type': 'scalar', 'host': 'local', + 'time': 3.0, 'value': 6, 'transform': 'rate', + 'summarize': 'p50'}, + ] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + summary = extract_result_metrics(data)[0]['short_json'][0] + + self.assertEqual(summary, { + 'name': 'custom.counter', 'type': 'scalar', 'host': 'local', + 'samples': 3, 'transform': 'rate', + 'summarize': 'p50', + 'summary': {'p50': 2.0}, + 'value_count': 2, + }) + + def test_producer_metadata_drives_total_summary(self): + metrics = [ + {'name': 'custom.drop.counter', 'type': 'scalar', 'host': 'local', + 'time': 1.0, 'value': 0, 'summarize': 'total'}, + {'name': 'custom.drop.counter', 'type': 'scalar', 'host': 'local', + 'time': 2.0, 'value': 3, 'summarize': 'total'}, + {'name': 'custom.drop.counter', 'type': 'scalar', 'host': 'local', + 'time': 3.0, 'value': 2, 'summarize': 'total'}, + ] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }], + }], + } + + summary = extract_result_metrics(data)[0]['short_json'][0] + + self.assertEqual(summary, { + 'name': 'custom.drop.counter', 'type': 'scalar', 'host': 'local', + 'samples': 3, 'summarize': 'total', + 'summary': {'total': 5}, 'value_count': 3, + }) + + def test_full_json_preserves_samples(self): + metric = {'name': 'throughput', 'type': 'scalar', 'time': 2, + 'value': 1, 'host': 'local', 'summarize': 'p50'} + sample = {'name': 'cpu.utilization', 'type': 'per-cpu', + 'host': 'local', 'time': 1, 'value': [99.5, None], + 'aggregation': 'busy-core-equivalents', + 'summarize': 'p90'} + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', + 'metrics': [metric, sample], + }], + }], + } + + rows = extract_result_metrics(data) + + self.assertEqual(len(rows), 1) + self.assertEqual(rows[0]['full_json'], [metric, sample]) + cpu_summary = rows[0]['short_json'][1] + self.assertEqual(cpu_summary, { + 'name': 'cpu.utilization', 'type': 'per-cpu', + 'host': 'local', 'samples': 1, 'value_count': 1, + 'aggregation': 'busy-core-equivalents', 'summarize': 'p90', + 'summary': {'p90': 0.995}, + 'cpus': 2, + }) + + def test_unusable_metrics_skip_their_attempt(self): + invalid = { + 'percentile': [{'name': 'x', 'value': 1, 'summarize': 'p150'}], + 'not-finite': [{'name': 'x', 'value': float('nan')}], + 'not-an-object': [1], + 'no-name': [{'value': 1}], + } + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [ + {'test': name, 'result': 'pass', 'metrics': metrics} + for name, metrics in invalid.items() + ] + [{ + 'test': 'valid', 'result': 'pass', + 'metrics': [{'name': 'x', 'value': 1, + 'summarize': 'p100'}], + }], + }], + } + + rows = extract_result_metrics(data) + + self.assertEqual([row['subtest'] for row in rows], ['valid']) + self.assertEqual(rows[0]['short_json'][0]['summary'], {'p100': 1}) + + def test_values_which_are_not_numbers_are_left_out(self): + metrics = [ + {'name': 'throughput', 'goal': 'higher', 'value': None}, + {'name': 'drops', 'goal': 'zero', 'value': True}, + {'name': 'latency', 'goal': 'lower', 'value': 'fast'}, + {'name': 'cpu', 'type': 'per-cpu', 'time': 1.0, + 'value': [1.0, 'busy']}, + {'name': 'rate', 'goal': 'higher', 'value': 10}, + ] + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', 'metrics': metrics, + }, { + 'test': 'no-numbers', 'result': 'pass', + 'metrics': [{'name': 'x', 'goal': 'higher', + 'value': None}], + }], + }], + } + + rows = extract_result_metrics(data) + + self.assertEqual([row['subtest'] for row in rows], ['case']) + self.assertEqual(rows[0]['full_json'], metrics) + self.assertEqual(rows[0]['short_json'], [{ + 'name': 'rate', 'goal': 'higher', 'samples': 1, + 'summarize': 'last', 'summary': {'last': 10}, + 'regression_value': 10, 'value_count': 1, + }]) + + def test_shared_case_names_store_one_row_per_attempt(self): + data = { + 'results': [{ + 'group': 'group', 'test': 'test', + 'results': [{ + 'test': 'case', 'result': 'pass', + 'metrics': [{'name': 'x', 'value': value}], + } for value in (1, 2)], + }], + } + + rows = extract_result_metrics(data) + + self.assertEqual(len(rows), 1) + self.assertEqual(rows[0]['full_json'], [{'name': 'x', 'value': 1}]) + + +if __name__ == '__main__': + unittest.main() From 3d2f9f883f07d9053078b99458dc801a22c2440c Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:20:57 -0700 Subject: [PATCH 06/18] contest: collector: store per-test performance metrics Add a result_metrics table with one row for each metric-bearing test-case attempt. Store both compact summaries and complete observation arrays so retries remain independent and details do not require one database row per sample. Teach the results collector to derive and upsert these records while ingesting a completed run. Log a failure to store them rather than stop ingesting results. Keep metrics out of results.json_full and the cached results, so /results?format=l2 responses and the combined results do not carry the observations. Signed-off-by: Stanislav Fomichev --- contest/results-collector.py | 62 +++++++++++++++++++++++++++++++++++- deploy/contest/db | 29 +++++++++++++++++ 2 files changed, 90 insertions(+), 1 deletion(-) diff --git a/contest/results-collector.py b/contest/results-collector.py index 2c6a9ff5..7db96741 100755 --- a/contest/results-collector.py +++ b/contest/results-collector.py @@ -8,9 +8,12 @@ import json import os import psycopg2 +from psycopg2.extras import Json, execute_values import requests import time +from backend.result_metrics import extract_result_metrics + """ Config: @@ -29,6 +32,7 @@ results-name=table-name wip-name=table-name branches-name=table-name +result-metrics-name=table-name """ def result_flatten(full): @@ -57,6 +61,16 @@ def result_flatten(full): return flat +def result_strip_metrics(full): + """Return a copy of a result without the metrics stored in result_metrics.""" + full = copy.deepcopy(full) + for test in full.get("results") or []: + for case in test.get("results") or []: + case.pop("metrics", None) + case.pop("retry_metrics", None) + return full + + class FetcherState: def __init__(self): self.config = configparser.ConfigParser() @@ -69,6 +83,8 @@ def __init__(self): self.tbl_res = self.config.get("db", "results-name", fallback="results") self.tbl_wip = self.config.get("db", "wip-name", fallback="results_pending") self.tbl_brn = self.config.get("db", "branches-name", fallback="branches") + self.tbl_met = self.config.get("db", "result-metrics-name", + fallback="result_metrics") db_name = self.config.get("db", "db") self.psql_conn = psycopg2.connect(database=db_name) @@ -123,9 +139,44 @@ def insert_result_psql(self, data): print(f"ERROR: {type(e).__module__}.{type(e).__name__}: {e.diag.message_primary}") print(f"ERROR: DETAIL: {e.diag.message_detail}") + def psql_insert_metrics(self, data): + rows = extract_result_metrics(data) + if not rows: + return + + values = [ + (data["branch"], data["branch"][-17:], data["remote"], + data["executor"], data["start"], row["group"], row["test"], + row["subtest"], Json(row["short_json"]), + Json(row["full_json"]), row["attempt"], row["result"], + row["link"]) + for row in rows + ] + fields = """( + branch, branch_date, remote, executor, t_start, + grp, test, subtest, short_json, full_json, + attempt, result, link + )""" + conflict = """( + branch, remote, executor, grp, test, subtest, attempt + )""" + with self.psql_conn.cursor() as cur: + execute_values( + cur, + f"""INSERT INTO {self.tbl_met} {fields} VALUES %s + ON CONFLICT {conflict} DO UPDATE SET + branch_date = EXCLUDED.branch_date, + t_start = EXCLUDED.t_start, + short_json = EXCLUDED.short_json, + full_json = EXCLUDED.full_json, + result = EXCLUDED.result, + link = EXCLUDED.link""", + values) + def psql_json_split(self, data): # return "normal" and "full" as json string or None # "full" will be None if they are the same to save storage + data = result_strip_metrics(data) full_s = json.dumps(data) if data.get("results") is None: # WIP result return full_s, None @@ -264,6 +315,14 @@ def insert_real(self, remote, run): self.psql_clear_wip(remote, run) self.insert_result_psql(data) + # Metrics come from the tests, which may print anything. The result + # is already stored, a failure here must not stop ingestion. + try: + self.psql_insert_metrics(data) + except Exception as e: # pylint: disable=broad-except + print(f"ERROR: failed to store metrics of {data['remote']}/" + f"{data['executor']} {data['branch']}: " + f"{type(e).__module__}.{type(e).__name__}: {e}") def write_json_atomic(path, data): @@ -286,7 +345,8 @@ def fetch_remote_run(fetcher, remote, run_info, remote_state): file = os.path.join(remote_state['dir'], os.path.basename(run_info['url'])) with open(file, "w") as fp: - json.dump(data, fp) + # The combined results are built from these, keep them compact + json.dump(result_strip_metrics(data), fp) return True diff --git a/deploy/contest/db b/deploy/contest/db index 6f86c34d..ae7ea80b 100644 --- a/deploy/contest/db +++ b/deploy/contest/db @@ -82,6 +82,35 @@ CREATE TABLE metrics ( value double precision ); +CREATE TABLE result_metrics ( + id bigserial PRIMARY KEY, + branch varchar(80) NOT NULL, + branch_date varchar(17) NOT NULL, + remote varchar(80) NOT NULL, + executor varchar(80) NOT NULL, + t_start timestamp NOT NULL, + grp varchar(80) NOT NULL, + test varchar(128) NOT NULL, + subtest varchar(256) NOT NULL, + short_json jsonb NOT NULL + CHECK (jsonb_typeof(short_json) = 'array'), + full_json jsonb NOT NULL + CHECK (jsonb_typeof(full_json) = 'array'), + attempt smallint NOT NULL DEFAULT 0 + CHECK (attempt IN (0, 1)), + result varchar(8) NOT NULL, + link text +); + +CREATE UNIQUE INDEX result_metrics_unique ON result_metrics + (branch, remote, executor, grp, test, subtest, attempt); +CREATE INDEX result_metrics_series ON result_metrics + (remote, executor, grp, test, subtest, branch_date DESC); + +GRANT SELECT, INSERT, UPDATE, DELETE ON result_metrics TO "nipa"; +GRANT USAGE, SELECT ON SEQUENCE result_metrics_id_seq TO "nipa"; +GRANT SELECT ON result_metrics TO "flask"; + CREATE TABLE stability ( remote varchar(80), executor varchar(80), From dcf184a62d616552ff909a2e252f97e539b87a5a Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:21:02 -0700 Subject: [PATCH 07/18] contest: backend: expose metric history and regressions Add APIs for compact metric history and complete observations. Compare tracked values with the median of up to five prior successful results with matching branch family, runner, test case, metric dimensions, and policy. Leave values which are not numbers out of the history. Return the latest results of each test case, and read five older ones as the baseline of the oldest returned, so a result is compared with the same history in the list and on its details page. Signed-off-by: Stanislav Fomichev --- contest/backend/query.py | 162 +++++++++++++++++++++++++- contest/backend/result_metrics.py | 136 ++++++++++++++++++++++ contest/tests/test_result_metrics.py | 163 ++++++++++++++++++++++++++- 3 files changed, 458 insertions(+), 3 deletions(-) diff --git a/contest/backend/query.py b/contest/backend/query.py index eb39daf3..b277d9bf 100644 --- a/contest/backend/query.py +++ b/contest/backend/query.py @@ -11,6 +11,10 @@ import re import datetime +from result_metrics import BRANCH_DATE_SUFFIX, REGRESSION_WINDOW, \ + branch_pattern, branch_prefix, evaluate_recent_regressions, \ + evaluate_regressions + app = Flask("NIPA contest query") @@ -20,6 +24,9 @@ # How many branches to query to get flakes for last month flake_cnt = 300 +# How many of the latest results of each test case to include in metric +# history +metric_history_limit = 50 @app.route('/') @@ -203,6 +210,158 @@ def remotes(): return rows +def _metric_start(timestamp): + start = timestamp.isoformat() + if timestamp.tzinfo is None: + start += '+00:00' + return start + + +def _metric_response(row, include_full=False): + data = { + 'id': row[0], + 'branch': row[1], + 'branch_date': row[2], + 'remote': row[3], + 'executor': row[4], + 'start': _metric_start(row[5]), + 'group': row[6], + 'test': row[7], + 'subtest': row[8], + 'short_json': row[9], + } + offset = 0 + if include_full: + data['full_json'] = row[10] + offset = 1 + data.update({ + 'attempt': row[10 + offset], + 'result': row[11 + offset], + 'link': row[12 + offset], + }) + return data + + +@app.route('/metrics') +def metrics(): + """Return compact summaries of the latest successful test-case attempts.""" + t1 = datetime.datetime.now() + + clauses = ["result = 'pass'"] + params = [] + filters = { + 'remote': 'remote', + 'executor': 'executor', + 'group': 'grp', + 'test': 'test', + 'subtest': 'subtest', + } + for arg, column in filters.items(): + value = request.args.get(arg) + if value: + clauses.append(f"{column} = %s") + params.append(value) + + requested_prefix = request.args.get('br-pfx') + if requested_prefix: + clauses.append("branch ~ %s") + params.append(branch_pattern(requested_prefix)) + + where = " WHERE " + " AND ".join(clauses) + with psql.cursor() as cur: + # Number the latest successful attempt of each run from the latest + # run, within each test case of a branch family and runner + cur.execute( + f"""WITH successful_metrics AS ( + SELECT DISTINCT ON ( + branch, remote, executor, grp, test, subtest) + id, branch, branch_date, remote, executor, t_start, + grp, test, subtest, short_json, + attempt, result, link + FROM result_metrics{where} + ORDER BY branch, remote, executor, grp, test, subtest, + attempt DESC + ), numbered_metrics AS ( + SELECT *, ROW_NUMBER() OVER ( + PARTITION BY regexp_replace(branch, %s, ''), + remote, executor, grp, test, + subtest + ORDER BY t_start DESC, id DESC) AS position + FROM successful_metrics + ) + SELECT id, branch, branch_date, remote, executor, t_start, + grp, test, subtest, short_json, attempt, result, link + FROM numbered_metrics + WHERE position <= %s + ORDER BY t_start, id""", + params + [BRANCH_DATE_SUFFIX, + metric_history_limit + REGRESSION_WINDOW]) + rows = cur.fetchall() + + # The results beyond the limit are only the baseline of the oldest ones + data = evaluate_recent_regressions( + [_metric_response(row) for row in rows], metric_history_limit) + print(f"Query for {len(data)} metric-bearing test-case results, up to " + f"{metric_history_limit} of each test case, took: " + f"{str(datetime.datetime.now() - t1)}") + return data + + +@app.route('/metrics/') +def metric_details(metric_id): + """Return all observations for one test-case attempt.""" + t1 = datetime.datetime.now() + with psql.cursor() as cur: + cur.execute( + """SELECT id, branch, branch_date, remote, executor, t_start, + grp, test, subtest, short_json, full_json, + attempt, result, link + FROM result_metrics + WHERE id = %s""", + (metric_id,)) + row = cur.fetchone() + + if row is None: + return {'error': 'metric result not found'}, 404 + + data = _metric_response(row, include_full=True) + if data['result'] == 'pass': + # The baseline of /metrics: the latest successful attempts of the + # previous runs of the test case + pattern = branch_pattern(branch_prefix(data['branch'])) + with psql.cursor() as cur: + cur.execute( + """SELECT id, branch, branch_date, remote, executor, t_start, + grp, test, subtest, short_json, + attempt, result, link + FROM (SELECT DISTINCT ON (branch) + id, branch, branch_date, remote, executor, + t_start, grp, test, subtest, short_json, + attempt, result, link + FROM result_metrics + WHERE result = 'pass' AND branch ~ %s + AND branch <> %s + AND remote = %s AND executor = %s + AND grp = %s AND test = %s AND subtest = %s + AND t_start <= %s + ORDER BY branch, attempt DESC) AS previous_runs + ORDER BY t_start DESC, id DESC + LIMIT %s""", + (pattern, data['branch'], data['remote'], data['executor'], + data['group'], data['test'], data['subtest'], + data['start'], REGRESSION_WINDOW)) + history = [_metric_response(item) + for item in reversed(cur.fetchall())] + current = {key: value for key, value in data.items() + if key != 'full_json'} + data['short_json'] = evaluate_regressions( + history + [current])[-1]['short_json'] + + print(f"Query for metric result {metric_id} took: " + f"{str(datetime.datetime.now() - t1)}") + return data + + @app.route('/stability') def stability(): # auto = query only tests which NIPA ignores based on stability @@ -300,8 +459,7 @@ def flaky_tests(): for row in rows: rem, exe, test, branch, br_date = row if group_pfx: - # Strip the trailing date stamp ("-YYYY-MM-DD--HH-MM", 18 chars) - key = (rem, exe, test, branch[:-18]) + key = (rem, exe, test, branch_prefix(branch)) else: key = (rem, exe, test) if not month: diff --git a/contest/backend/result_metrics.py b/contest/backend/result_metrics.py index e24205ff..1f8ce7be 100644 --- a/contest/backend/result_metrics.py +++ b/contest/backend/result_metrics.py @@ -2,8 +2,12 @@ """Helpers for extracting and summarizing metrics from a NIPA result.""" +import collections +import copy import json import math +import re +import statistics def _series_key(metric): @@ -193,6 +197,138 @@ def summarize_metrics(metrics): return summaries +_BRANCH_DATE = r'[0-9]{4}-[0-9]{2}-[0-9]{2}--[0-9]{2}-[0-9]{2}' +# Date stamp of NIPA branches, valid in Python and PostgreSQL regexes +BRANCH_DATE_SUFFIX = rf'-{_BRANCH_DATE}$' +_BRANCH_DATE_SUFFIX_RE = re.compile(BRANCH_DATE_SUFFIX) + + +def branch_prefix(branch): + return _BRANCH_DATE_SUFFIX_RE.sub('', branch) + + +def branch_pattern(prefix): + return rf'^{re.escape(prefix)}-{_BRANCH_DATE}$' + + +def _regression_series_key(row, metric): + dimensions = { + key: value for key, value in metric.items() + if key not in { + 'summary', 'samples', 'value_count', 'cpus', 'display_range', + 'display_scale', 'display_label', 'regression_value', + 'regression_result'} + } + identity = { + 'branch_prefix': branch_prefix(row['branch']), + 'remote': row['remote'], + 'executor': row['executor'], + 'group': row['group'], + 'test': row['test'], + 'subtest': row['subtest'], + 'metric': dimensions, + } + return json.dumps(identity, sort_keys=True, separators=(',', ':')) + + +# How much worse than the median of its recent results a performance metric +# has to get to regress +REGRESSION_TOLERANCE = 0.1 +# How many earlier results the median is taken of +REGRESSION_WINDOW = 5 + + +def evaluate_regressions(rows, *, window=REGRESSION_WINDOW): + """Annotate compact metric rows using earlier comparable results. + + The goal of a metric gives its direction: an error metric must be zero, + a performance metric regresses when it is worse than the median of its + recent results by more than REGRESSION_TOLERANCE. NIPA owns the + historical cohort: successful results from the same branch family, + runner, test case, and metric series. Input ordering is preserved in the + returned deep copy. + """ + evaluated = copy.deepcopy(rows) + history = {} + ordered = sorted(enumerate(evaluated), + key=lambda item: (item[1]['start'], item[0])) + + for _index, row in ordered: + if row['result'] != 'pass': + continue + for metric in row['short_json']: + goal = metric.get('goal') + if goal not in ('higher', 'lower', 'zero'): + continue + + try: + current = metric['regression_value'] + # Anything else in the history would stop all later checks + if not _is_number(current): + continue + if goal == 'zero': + metric['regression_result'] = { + 'status': 'error' if current else 'ok', + 'current': _summary_number(current), + } + continue + key = _regression_series_key(row, metric) + baseline_values = history.setdefault(key, [])[-window:] + result = { + 'status': 'insufficient-data', + 'current': _summary_number(current), + 'baseline_samples': len(baseline_values), + } + if baseline_values: + baseline = statistics.median(baseline_values) + threshold = REGRESSION_TOLERANCE * abs(baseline) + if goal == 'higher': + worsening = baseline - current + else: + worsening = current - baseline + result.update({ + 'status': ('regression' + if worsening > threshold else 'ok'), + 'baseline': _summary_number(baseline), + 'threshold': _summary_number(threshold), + }) + if baseline: + result['delta_percent'] = _summary_number( + (current - baseline) / abs(baseline) * 100) + metric['regression_result'] = result + history[key].append(current) + except (KeyError, TypeError, ValueError, ArithmeticError): + continue + + return evaluated + + +def _case_key(row): + return (branch_prefix(row['branch']), row['remote'], row['executor'], + row['group'], row['test'], row['subtest']) + + +def evaluate_recent_regressions(rows, limit, *, window=REGRESSION_WINDOW): + """Evaluate rows, then return the latest `limit` results of each test case. + + The older rows only serve as the baseline: given `window` of them for + each test case, even the oldest returned result is compared with a full + window of history. + """ + evaluated = evaluate_regressions(rows, window=window) + latest = sorted(enumerate(evaluated), + key=lambda item: (item[1]['start'], item[0]), + reverse=True) + counts = collections.Counter() + keep = set() + for index, row in latest: + key = _case_key(row) + counts[key] += 1 + if counts[key] <= limit: + keep.add(index) + return [row for index, row in enumerate(evaluated) if index in keep] + + def extract_result_metrics(data): """Return one summarized/full metric record per test-case attempt.""" rows = [] diff --git a/contest/tests/test_result_metrics.py b/contest/tests/test_result_metrics.py index 064fd639..684643d4 100644 --- a/contest/tests/test_result_metrics.py +++ b/contest/tests/test_result_metrics.py @@ -2,7 +2,8 @@ import unittest -from contest.backend.result_metrics import extract_result_metrics +from contest.backend.result_metrics import evaluate_recent_regressions, \ + evaluate_regressions, extract_result_metrics class TestResultMetrics(unittest.TestCase): @@ -298,6 +299,133 @@ def test_producer_metadata_drives_total_summary(self): 'summary': {'total': 5}, 'value_count': 3, }) + def test_regressions_use_previous_comparable_results(self): + rows = [] + for index, value in enumerate([100, 88, 100, 100, 100, 100, 88]): + rows.append({ + 'id': index + 1, + 'branch': f'net-next-2026-01-{index + 1:02d}--00-00', + 'remote': 'metal', 'executor': 'executor', + 'start': f'2026-01-{index + 1:02d}T00:00:00+00:00', + 'group': 'group', 'test': 'test', 'subtest': 'case', + 'attempt': 0, 'result': 'pass', + 'short_json': [{ + 'name': 'throughput', 'goal': 'higher', + 'host': 'local', 'summarize': 'last', + 'samples': 1, 'value_count': 1, + 'summary': {'last': value}, + 'regression_value': value, + }], + }) + + evaluated = evaluate_regressions(rows) + + self.assertNotIn('regression_result', rows[-1]['short_json'][0]) + first = evaluated[0]['short_json'][0]['regression_result'] + self.assertEqual(first, { + 'status': 'insufficient-data', 'current': 100, + 'baseline_samples': 0, + }) + early = evaluated[1]['short_json'][0]['regression_result'] + self.assertEqual(early, { + 'status': 'regression', 'current': 88, + 'baseline_samples': 1, 'baseline': 100, + 'threshold': 10.0, 'delta_percent': -12.0, + }) + stable = evaluated[5]['short_json'][0]['regression_result'] + self.assertEqual(stable['status'], 'ok') + self.assertEqual(stable['baseline'], 100) + regressed = evaluated[6]['short_json'][0]['regression_result'] + self.assertEqual(regressed, { + 'status': 'regression', 'current': 88, + 'baseline_samples': 5, 'baseline': 100, + 'threshold': 10.0, 'delta_percent': -12.0, + }) + + other_runner = [dict(rows[-1], id=20, executor='other')] + isolated = evaluate_regressions(rows[:-1] + other_runner)[-1] + self.assertEqual( + isolated['short_json'][0]['regression_result']['status'], + 'insufficient-data') + + def test_error_metrics_must_be_zero(self): + rows = [{ + 'id': index + 1, + 'branch': f'net-next-2026-01-{index + 1:02d}--00-00', + 'remote': 'metal', 'executor': 'executor', + 'start': f'2026-01-{index + 1:02d}T00:00:00+00:00', + 'group': 'group', 'test': 'test', 'subtest': 'case', + 'attempt': 0, 'result': 'pass', + 'short_json': [{ + 'name': 'ring.rx_alloc_fail', 'goal': 'zero', + 'summarize': 'last', 'samples': 1, 'value_count': 1, + 'summary': {'last': value}, 'regression_value': value, + }], + } for index, value in enumerate([0, 2])] + + evaluated = evaluate_regressions(rows) + + self.assertEqual(evaluated[0]['short_json'][0]['regression_result'], + {'status': 'ok', 'current': 0}) + self.assertEqual(evaluated[1]['short_json'][0]['regression_result'], + {'status': 'error', 'current': 2}) + + def test_values_which_are_not_numbers_stay_out_of_history(self): + rows = [{ + 'id': index + 1, + 'branch': f'net-next-2026-01-{index + 1:02d}--00-00', + 'remote': 'metal', 'executor': 'executor', + 'start': f'2026-01-{index + 1:02d}T00:00:00+00:00', + 'group': 'group', 'test': 'test', 'subtest': 'case', + 'attempt': 0, 'result': 'pass', + 'short_json': [{ + 'name': 'throughput', 'goal': 'higher', + 'summarize': 'last', 'samples': 1, 'value_count': 1, + 'summary': {'last': value}, 'regression_value': value, + }], + } for index, value in enumerate([None, 100, 'fast', 100, 50])] + + evaluated = evaluate_regressions(rows) + + results = [row['short_json'][0].get('regression_result') + for row in evaluated] + self.assertIsNone(results[0]) + self.assertEqual(results[1]['status'], 'insufficient-data') + self.assertIsNone(results[2]) + self.assertEqual(results[3]['status'], 'ok') + self.assertEqual(results[4]['status'], 'regression') + self.assertEqual(results[4]['baseline_samples'], 2) + + def test_malformed_regression_does_not_stop_evaluation(self): + rows = [] + for index in range(2): + rows.append({ + 'id': index + 1, + 'branch': f'net-next-2026-01-{index + 1:02d}--00-00', + 'remote': 'metal', 'executor': 'executor', + 'start': f'2026-01-{index + 1:02d}T00:00:00+00:00', + 'group': 'group', 'test': 'test', 'subtest': 'case', + 'attempt': 0, 'result': 'pass', + 'short_json': [{ + 'name': 'invalid', 'goal': 'higher', + 'summarize': 'last', 'samples': 1, 'value_count': 1, + 'summary': {'last': 10}, + }, { + 'name': 'valid', 'goal': 'higher', + 'summarize': 'last', 'samples': 1, 'value_count': 1, + 'summary': {'last': 20}, + 'regression_value': 20, + }], + }) + + evaluated = evaluate_regressions(rows) + + self.assertNotIn( + 'regression_result', evaluated[1]['short_json'][0]) + self.assertEqual( + evaluated[1]['short_json'][1]['regression_result']['status'], + 'ok') + def test_full_json_preserves_samples(self): metric = {'name': 'throughput', 'type': 'scalar', 'time': 2, 'value': 1, 'host': 'local', 'summarize': 'p50'} @@ -402,6 +530,39 @@ def test_shared_case_names_store_one_row_per_attempt(self): self.assertEqual(len(rows), 1) self.assertEqual(rows[0]['full_json'], [{'name': 'x', 'value': 1}]) + def test_recent_regressions_keep_the_baseline_of_the_oldest(self): + def row(index, executor, value): + return { + 'id': index, + 'branch': f'net-next-2026-01-{index:02d}--00-00', + 'remote': 'metal', 'executor': executor, + 'start': f'2026-01-{index:02d}T00:00:00+00:00', + 'group': 'group', 'test': 'test', 'subtest': 'case', + 'attempt': 0, 'result': 'pass', + 'short_json': [{ + 'name': 'throughput', 'goal': 'higher', + 'summarize': 'last', 'samples': 1, 'value_count': 1, + 'summary': {'last': value}, 'regression_value': value, + }], + } + + # Nine results of one runner, the latest three of them regressed, + # and one result of another runner + rows = [row(index, 'executor', 100 if index < 7 else 80) + for index in range(1, 10)] + rows.append(row(10, 'other', 80)) + + recent = evaluate_recent_regressions(rows, 3) + + self.assertEqual([entry['id'] for entry in recent], [7, 8, 9, 10]) + results = [entry['short_json'][0]['regression_result'] + for entry in recent] + self.assertEqual(results[0], { + 'status': 'regression', 'current': 80, 'baseline_samples': 5, + 'baseline': 100, 'threshold': 10.0, 'delta_percent': -20.0, + }) + self.assertEqual(results[3]['status'], 'insufficient-data') + if __name__ == '__main__': unittest.main() From eb48a0bb717158b6a9cca902ce86bff925265691 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:21:09 -0700 Subject: [PATCH 08/18] contest: add performance metric API fixtures Add compact history and full detail fixtures for devmem and iperf metrics. Trim two hardware runs to the eight busiest CPUs per host and every second system monitor sample. Check that stored summaries and regression results follow from the observations. Signed-off-by: Stanislav Fomichev --- contest/tests/test_result_metrics.py | 34 +- ui/mock/result-metric-details.json | 1888 ++++++++++++++++++++++++++ ui/mock/result-metrics.json | 1184 ++++++++++++++++ 3 files changed, 3105 insertions(+), 1 deletion(-) create mode 100644 ui/mock/result-metric-details.json create mode 100644 ui/mock/result-metrics.json diff --git a/contest/tests/test_result_metrics.py b/contest/tests/test_result_metrics.py index 684643d4..31527248 100644 --- a/contest/tests/test_result_metrics.py +++ b/contest/tests/test_result_metrics.py @@ -1,9 +1,25 @@ # SPDX-License-Identifier: GPL-2.0 +import json +import os import unittest from contest.backend.result_metrics import evaluate_recent_regressions, \ - evaluate_regressions, extract_result_metrics + evaluate_regressions, extract_result_metrics, summarize_metrics + + +UI_MOCK_DIR = os.path.join(os.path.dirname(__file__), '..', '..', 'ui', 'mock') + + +def _load_ui_mock(name): + with open(os.path.join(UI_MOCK_DIR, name), encoding='utf-8') as fp: + return json.load(fp) + + +def _without_full_json(details): + return [{key: value for key, value in detail.items() + if key != 'full_json'} + for detail in details] class TestResultMetrics(unittest.TestCase): @@ -426,6 +442,22 @@ def test_malformed_regression_does_not_stop_evaluation(self): evaluated[1]['short_json'][1]['regression_result']['status'], 'ok') + def test_ui_mock_details_match_overview(self): + rows = _load_ui_mock('result-metrics.json') + details = _load_ui_mock('result-metric-details.json') + + self.assertEqual(_without_full_json(details), rows) + + def test_ui_mock_summaries_match_observations(self): + """Fixture summaries and regressions follow from the observations.""" + details = _load_ui_mock('result-metric-details.json') + rows = _without_full_json(details) + for row, detail in zip(rows, details): + row['short_json'] = summarize_metrics(detail['full_json']) + + self.assertEqual(evaluate_regressions(rows), + _without_full_json(details)) + def test_full_json_preserves_samples(self): metric = {'name': 'throughput', 'type': 'scalar', 'time': 2, 'value': 1, 'host': 'local', 'summarize': 'p50'} diff --git a/ui/mock/result-metric-details.json b/ui/mock/result-metric-details.json new file mode 100644 index 00000000..4a5281f7 --- /dev/null +++ b/ui/mock/result-metric-details.json @@ -0,0 +1,1888 @@ +[ + { + "group": "selftests-drivers-net-perf", + "test": "devmem-py", + "subtest": "devmem-kperf", + "link": "/artifacts/0/test-outputs/0-devmem-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 215.378 + }, + "regression_value": 215.378, + "value_count": 1, + "regression_result": { + "status": "insufficient-data", + "current": 215.378, + "baseline_samples": 0 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.079788 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 3.050889 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 2.148583 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 633.769492 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 5.486676 + }, + "value_count": 17 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.054877 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.573283 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.762722 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 1.29674 + }, + "value_count": 17 + } + ], + "attempt": 0, + "result": "pass", + "id": 1, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-33", + "branch_date": "2026-09-25--21-33", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:33:21.292718+00:00", + "full_json": [ + {"goal":"higher","name":"throughput","value":215.37800000000001,"origin":"selftest"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.002457,"value":[0.01,0.049996881441329605,0.0,0.01999937628803309,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.002457,"value":[0.0,0.03999937628832413,0.02999937628832413,0.0499981288658455,0.009999376288033091,0.0,0.01,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.002457,"value":[0.13999501030892134,0.0,0.0,0.01999937628803309,0.0,0.01,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.002457,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.002457,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.002457,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.002457,"value":[8.0808,3.0,2.0,3.9604,0.0,0.9901,0.9901,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.00751,"value":[0.01,0.05999688144132961,0.0,0.09999937628803308,0.01,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.00751,"value":[0.01,0.03999937628832413,0.14999937628832413,0.3399981288658455,0.01999937628803309,0.0,0.01,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.00751,"value":[0.28999501030892133,0.0,0.01,0.02999937628803309,0.0,0.01,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.00751,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.00751,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.00751,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.00751,"value":[6.1224,0.0,2.9703,8.8235,1.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.014584,"value":[0.02,0.05999688144132961,0.01,0.4499993762880331,0.01,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.014584,"value":[0.01,0.03999937628832413,0.15999937628832414,1.6899981288658454,0.02999937628803309,0.0,0.02,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.014584,"value":[0.47999501030892133,0.0,0.01,0.039999376288033094,0.0,0.01,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.014584,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.014584,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.014584,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.014584,"value":[11.1111,0.0,2.0,93.0693,0.0,0.0,1.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.019084,"value":[0.02,0.05999688144132961,0.01,0.4599993762880331,0.01,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.019084,"value":[0.02,0.03999937628832413,0.19999937628832412,1.6999981288658454,0.02999937628803309,0.0,0.02,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.019084,"value":[0.6299950103089214,0.0,0.01,0.04999937628803309,0.0,0.04,0.0,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.019084,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.019084,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.019084,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.019084,"value":[6.1224,0.0,3.0,3.0,0.0,0.0,0.0,2.9412],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.024719,"value":[0.02,0.05999688144132961,0.01,0.4599993762880331,0.02,0.0,0.0,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.024719,"value":[0.02,0.03999937628832413,0.3999993762883241,1.7099981288658455,0.039999376288033094,0.01,0.02,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.024719,"value":[0.7499950103089214,0.0,0.01,0.05999937628803309,0.0,0.04,0.0,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.024719,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.024719,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.024719,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.024719,"value":[6.0606,0.0,0.0,1.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":12.030058,"value":[0.02,0.05999688144132961,0.01,0.46999937628803307,0.03,0.0,0.0,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":12.030058,"value":[0.03,0.04999937628832413,0.4099993762883241,1.7499981288658455,0.04999937628803309,0.01,0.03,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":12.030058,"value":[0.9699950103089213,0.0,0.01,0.06999937628803309,0.01,0.04,0.0,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":12.030058,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":12.030058,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":12.030058,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":12.030058,"value":[12.2449,0.0,1.0101,6.9307,2.9412,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":14.035363,"value":[0.02,0.06999688144132961,0.02,0.5799993762880331,0.04,0.0,0.0,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":14.035363,"value":[0.04,0.04999937628832413,0.4099993762883241,2.1199981288658454,0.05999937628803309,0.01,0.03,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":14.035363,"value":[1.2199950103089214,0.03,0.05,0.0799993762880331,0.01,0.08,0.0,0.07],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":14.035363,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":14.035363,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":14.035363,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":14.035363,"value":[12.7451,0.0,0.0,4.0,1.9802,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":16.040824,"value":[0.02,0.08999688144132961,0.03,0.5999993762880331,0.05,0.0,0.02,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":16.040824,"value":[0.04,0.9099993762883242,1.1099993762883242,2.1599981288658454,0.9799993762880331,0.01,0.79,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":16.040824,"value":[1.4599950103089214,0.05,0.06,0.08999937628803309,0.01,1.11,0.0,1.1],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":16.040824,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":16.040824,"value":1271.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":16.040824,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":16.040824,"value":[14.0,89.1089,72.0,4.9505,91.9192,100.0,77.0,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":18.046277,"value":[0.02,0.1099968814413296,0.04,0.5999993762880331,0.06,0.0,0.04,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":18.046277,"value":[0.05,2.599999376288324,1.2699993762883242,2.1999981288658454,2.809999376288033,0.01,2.29,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":18.046277,"value":[1.7099950103089214,0.07,0.07,0.09999937628803308,0.02,3.11,0.0,3.1],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":18.046277,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":18.046277,"value":1271.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":18.046277,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":18.046277,"value":[12.8713,90.625,0.0,3.0,93.0,100.0,76.7677,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":20.052674,"value":[0.03,0.1499968814413296,0.05,0.629999376288033,0.08,0.0,0.07,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":20.052674,"value":[0.05,4.289999376288324,1.4699993762883241,2.2499981288658457,4.639999376288033,0.01,3.77,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":20.052674,"value":[1.9499950103089214,0.09,0.08,0.09999937628803308,0.03,5.11,0.01,5.11],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":20.052674,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":20.052674,"value":1271.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":20.052674,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":20.052674,"value":[12.0,89.899,4.0,3.9604,93.8776,100.0,76.5306,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":22.058647,"value":[0.03,0.1799968814413296,0.07,0.639999376288033,0.09,0.0,0.09,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":22.058647,"value":[0.05,5.989999376288324,2.569999376288324,2.2799981288658455,6.4599993762880334,0.01,5.24,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":22.058647,"value":[2.199995010308921,0.11,0.1,0.1099993762880331,0.04,7.11,0.01,7.11],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":22.058647,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":22.058647,"value":1271.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":22.058647,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":22.058647,"value":[12.1212,90.0,74.2268,2.0,93.0,100.0,77.3196,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":24.064637,"value":[0.03,0.2099968814413296,0.08,0.7099993762880331,0.1,0.0,0.1,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":24.064637,"value":[0.06,7.7199993762883246,3.079999376288324,2.3399981288658456,8.319999376288033,0.01,6.6,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":24.064637,"value":[2.4599950103089214,0.13,0.11,0.11999937628803309,0.05,9.11,0.02,9.11],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":24.064637,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":24.064637,"value":1322.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":24.064637,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":24.064637,"value":[12.1212,89.899,24.2424,4.0,96.0,100.0,66.3265,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":26.070649,"value":[0.03,0.2299968814413296,0.1,0.7199993762880331,0.12,0.0,0.12,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":26.070649,"value":[0.06,9.419999376288324,4.379999376288324,2.3999981288658456,10.199999376288034,0.01,7.88,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":26.070649,"value":[2.699995010308921,0.15,0.13,0.1299993762880331,0.06,11.11,0.02,11.11],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":26.070649,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":26.070649,"value":1322.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":26.070649,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":26.070649,"value":[12.2449,89.6907,64.9485,4.9505,97.0,100.0,67.3469,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":28.075824,"value":[0.03,0.2599968814413296,0.12,0.7399993762880331,0.14,0.0,0.15,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":28.075824,"value":[0.07,11.099999376288324,5.309999376288324,2.4399981288658457,12.089999376288032,0.01,9.19,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":28.075824,"value":[2.949995010308921,0.17,0.15,0.1299993762880331,0.07,13.11,0.02,13.11],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":28.075824,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":28.075824,"value":1322.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":28.075824,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":28.075824,"value":[11.1111,89.899,52.0,3.0,96.0,100.0,67.6768,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":30.081377,"value":[0.03,0.2899968814413296,0.14,0.749999376288033,0.15,0.0,0.17,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":30.081377,"value":[0.07,12.719999376288325,6.209999376288324,2.4899981288658455,13.949999376288034,0.01,10.48,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":30.081377,"value":[3.179995010308921,0.19,0.16,0.13999937628803308,0.08,15.12,0.03,15.09],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":30.081377,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":30.081377,"value":1322.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":30.081377,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":30.081377,"value":[12.0,85.7143,73.4043,3.0303,95.0,100.0,68.6869,98.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":32.086681,"value":[0.09,0.2899968814413296,0.14,0.8499993762880331,0.15,0.0,0.17,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":32.086681,"value":[0.12,12.849999376288324,6.319999376288324,2.8599981288658456,13.989999376288033,0.01,10.49,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":32.086681,"value":[3.3399950103089213,0.19,0.17,0.1499993762880331,0.08,15.16,0.03,15.09],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":32.086681,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":32.086681,"value":1322.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":32.086681,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":32.086681,"value":[18.0,0.0,6.0606,26.0,0.0,0.9901,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":34.046139,"value":[0.09953899431158789,0.2899968814413296,0.15907798862317576,0.8690773649112089,0.16,0.0,0.17,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":34.046139,"value":[0.12953899431158789,12.849999376288324,6.5090773649115,2.898154106112197,13.989999376288033,0.01,10.5,0.03],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":34.046139,"value":[3.5049239477328955,0.21907798862317576,0.17,0.15953837059962098,0.08,15.179077988623176,0.03,15.118616982934764],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":34.046139,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":34.046139,"value":1322.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":34.046139,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":34.046139,"value":[12.7451,2.0202,3.9604,7.0,0.0,2.0,0.0,3.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.002457,"value":[0.01,0.00999937628832413,0.0,0.01999875257664826,0.039998128864972386,0.01,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.002457,"value":[0.0,0.00999937628832413,0.0,0.029998128864972388,0.029998128864972388,0.0,0.009999376288033091,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.002457,"value":[0.229992515463382,0.0,0.0,0.00999937628832413,0.0,0.01,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.002457,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.002457,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.002457,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.002457,"value":[11.8812,0.0,0.0,0.0,1.0,1.9608,0.9901,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.00751,"value":[0.01,0.00999937628832413,0.0,0.01999875257664826,0.039998128864972386,0.01,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.00751,"value":[0.0,0.00999937628832413,0.0,0.039998128864972386,0.039998128864972386,0.01,0.01999937628803309,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.00751,"value":[0.37999251546338203,0.0,0.0,0.00999937628832413,0.01,0.01,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.00751,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.00751,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.00751,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.00751,"value":[8.0,0.9901,0.0,0.0,0.0,1.0,0.9901,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.014584,"value":[0.01,0.00999937628832413,0.0,0.01999875257664826,0.039998128864972386,0.01,0.01,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.014584,"value":[0.01,0.00999937628832413,0.0,0.039998128864972386,0.039998128864972386,0.01,0.01999937628803309,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.014584,"value":[0.519992515463382,0.01,0.01,0.00999937628832413,0.01,0.01,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.014584,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.014584,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.014584,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.014584,"value":[7.1429,1.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.019084,"value":[0.01,0.02999937628832413,0.02,0.029998752576648257,0.039998128864972386,0.04,0.02,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.019084,"value":[0.02,0.019999376288324128,0.01,0.04999812886497239,0.04999812886497239,0.02,0.02999937628803309,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.019084,"value":[0.729992515463382,0.01,0.01,0.019999376288324128,0.01,0.02,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.019084,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.019084,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.019084,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.019084,"value":[10.8911,0.0,0.9901,0.0,0.0,0.9901,0.0,1.9802],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.024719,"value":[0.01,0.02999937628832413,0.03,0.029998752576648257,0.05999812886497238,0.05,0.02,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.024719,"value":[0.03,0.019999376288324128,0.01,0.05999812886497238,0.06999812886497239,0.02,0.02999937628803309,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.024719,"value":[0.839992515463382,0.01,0.01,0.019999376288324128,0.01,0.02,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.024719,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.024719,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.024719,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.024719,"value":[6.0606,0.0,0.9901,1.0,0.9901,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":12.030058,"value":[0.02,0.04999937628832413,0.04,0.029998752576648257,0.06999812886497239,0.05,0.02,0.04],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":12.030058,"value":[0.03,0.02999937628832413,0.02,0.06999812886497239,0.07999812886497239,0.02,0.039999376288033094,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":12.030058,"value":[0.999992515463382,0.02,0.01,0.019999376288324128,0.02,0.02,0.02,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":12.030058,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":12.030058,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":12.030058,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":12.030058,"value":[8.0808,2.0,1.9608,0.9901,1.0,0.0,1.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":14.035363,"value":[0.02,0.08999937628832413,0.04,0.08999875257664826,0.07999812886497239,0.06,0.21,0.04],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":14.035363,"value":[0.04,0.06999937628832413,0.02,0.07999812886497239,0.07999812886497239,0.03,0.3099993762880331,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":14.035363,"value":[1.249992515463382,0.02,0.01,0.02999937628832413,0.02,0.02,0.02,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":14.035363,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":14.035363,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":14.035363,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":14.035363,"value":[13.8614,4.0404,0.0,6.0,0.0,1.0,44.4444,1.9802],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":16.040824,"value":[0.02,0.09999937628832413,0.04,0.08999875257664826,0.07999812886497239,0.06,0.21,0.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":16.040824,"value":[0.04,0.07999937628832413,0.13,0.08999812886497238,0.24999812886497239,0.03,0.40999937628803307,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":16.040824,"value":[1.399992515463382,0.18,0.01,0.14999937628832413,0.02,0.22,0.15,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":16.040824,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":16.040824,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":16.040824,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":16.040824,"value":[9.0909,19.6078,10.0,13.8614,16.0,21.0,24.2424,2.9412],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":18.046277,"value":[0.03,0.09999937628832413,0.04,0.09999875257664825,0.08999812886497238,0.06,0.21,0.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":18.046277,"value":[0.05,0.07999937628832413,0.31,0.08999812886497238,0.5699981288649724,0.04,0.5699993762880331,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":18.046277,"value":[1.669992515463382,0.53,0.01,0.3999993762883241,0.03,0.59,0.42,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":18.046277,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":18.046277,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":18.046277,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":18.046277,"value":[15.8416,19.0,10.0,14.2857,18.6275,21.5686,24.2424,0.9901],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":20.052674,"value":[0.03,0.09999937628832413,0.04,0.09999875257664825,0.08999812886497238,0.06,0.21,0.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":20.052674,"value":[0.05,0.07999937628832413,0.54,0.09999812886497239,0.9399981288649724,0.04,0.7299993762880331,0.27],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":20.052674,"value":[1.9099925154633821,0.89,0.02,0.6599993762883242,0.04,0.97,0.7,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":20.052674,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":20.052674,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":20.052674,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":20.052674,"value":[10.0,19.1919,10.2041,14.8515,19.0,20.202,25.0,11.2245],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":22.058647,"value":[0.03,0.09999937628832413,0.04,0.09999875257664825,0.08999812886497238,0.06,0.21,0.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":22.058647,"value":[0.06,0.07999937628832413,0.79,0.09999812886497239,1.2899981288649724,0.04,0.9299993762880331,0.65],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":22.058647,"value":[2.139992515463382,1.24,0.02,0.9199993762883242,0.06,1.36,0.98,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":22.058647,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":22.058647,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":22.058647,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":22.058647,"value":[12.1212,19.802,12.8713,14.1414,20.0,20.7921,27.2727,19.802],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":24.064637,"value":[0.04,0.09999937628832413,0.04,0.09999875257664825,0.08999812886497238,0.06,0.22,0.07],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":24.064637,"value":[0.07,0.07999937628832413,1.01,0.09999812886497239,1.6499981288649723,0.05,1.0899993762880331,1.04],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":24.064637,"value":[2.369992515463382,1.56,0.02,1.189999376288324,0.08,1.76,1.25,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":24.064637,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":24.064637,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":24.064637,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":24.064637,"value":[12.0,19.0,12.1212,15.0,18.0,22.7723,22.449,6.8627],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":26.070649,"value":[0.04,0.09999937628832413,0.04,0.09999875257664825,0.08999812886497238,0.06,0.22,0.08],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":26.070649,"value":[0.07,0.08999937628832413,1.24,0.09999812886497239,2.0099981288649724,0.05,1.239999376288033,1.04],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":26.070649,"value":[2.569992515463382,1.92,0.03,1.459999376288324,0.09,2.18,1.49,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":26.070649,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":26.070649,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":26.070649,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":26.070649,"value":[11.1111,19.0,11.1111,15.0,19.802,22.0,21.6495,1.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":28.075824,"value":[0.04,0.09999937628832413,0.04,0.10999875257664826,0.08999812886497238,0.06,0.22,0.08],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":28.075824,"value":[0.08,0.08999937628832413,1.48,0.10999812886497239,2.3999981288649725,0.05,1.3999993762880332,1.05],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":28.075824,"value":[2.779992515463382,2.28,0.03,1.7299993762883241,0.1,2.59,1.75,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":28.075824,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":28.075824,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":28.075824,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":28.075824,"value":[12.0,19.1919,12.1212,14.2857,20.0,21.2121,24.2424,1.0101],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":30.081377,"value":[0.04,0.09999937628832413,0.04,0.10999875257664826,0.08999812886497238,0.07,0.22,0.08],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":30.081377,"value":[0.09,0.08999937628832413,1.7,0.10999812886497239,2.729998128864972,0.05,1.539999376288033,1.33],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":30.081377,"value":[2.939992515463382,2.63,0.03,1.9999993762883241,0.11,2.97,2.0,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":30.081377,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":30.081377,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":30.081377,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":30.081377,"value":[9.0909,18.1818,11.0,14.0,17.0,19.802,22.449,17.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":32.086681,"value":[0.05,0.10999937628832412,0.05,0.10999875257664826,0.08999812886497238,0.07,0.22,0.08],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":32.086681,"value":[0.1,0.09999937628832413,1.71,0.11999812886497238,2.729998128864972,0.05,1.6799993762880332,1.34],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":32.086681,"value":[3.129992515463382,2.64,0.04,2.009999376288324,0.11,2.98,2.01,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":32.086681,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":32.086681,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":32.086681,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":32.086681,"value":[11.0,1.9802,0.0,1.9802,0.0,1.0,1.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":34.046139,"value":[0.05,0.10999937628832412,0.05,0.11999875257664826,0.09999812886497239,0.07,0.22,0.08],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":34.046139,"value":[0.10953899431158788,0.09999937628832413,1.719538994311588,0.1399981288649724,2.729998128864972,0.05,1.6799993762880332,1.34],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":34.046139,"value":[3.315843464266509,2.64,0.04,2.009999376288324,0.11,2.98,2.01,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":34.046139,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":34.046139,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":34.046139,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":34.046139,"value":[11.0,0.0,0.9901,0.0,0.0,0.0,0.0,1.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"} + ] + }, + { + "group": "selftests-drivers-net-perf", + "test": "iperf-py", + "subtest": "iperf-tcp", + "link": "/artifacts/0/test-outputs/1-iperf-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 57.294036 + }, + "regression_value": 57.294036, + "value_count": 1, + "regression_result": { + "status": "insufficient-data", + "current": 57.294036, + "baseline_samples": 0 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.074759 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 1.554998 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.827338 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 2.680909 + }, + "value_count": 6 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.049845 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.737702 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.269162 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 1.015569 + }, + "value_count": 6 + } + ], + "attempt": 0, + "result": "pass", + "id": 2, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-33", + "branch_date": "2026-09-25--21-33", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:33:21.292718+00:00", + "full_json": [ + {"goal":"higher","name":"throughput","value":57.29403584064005,"origin":"selftest"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.002449,"value":[0.0,0.0,0.01,0.009998450218699873,0.01,0.009998450218408833,0.009998450218408833,0.03999380087392637],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.002449,"value":[0.01,0.0,0.52,0.039998450218699876,0.04,0.0,0.0,0.1599752034957055],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.002449,"value":[0.17998605196364223,0.38,0.0,0.009998450218699873,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.002449,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.002449,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.002449,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.002449,"value":[10.8911,35.8491,53.4653,3.0303,4.9505,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.008102,"value":[0.0,0.0,0.02,0.01999845021869987,0.01,0.029998450218408836,0.019998450218408834,0.03999380087392637],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.008102,"value":[0.01,0.0,1.53,0.09999845021869987,0.04,0.0,0.94,0.1599752034957055],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.008102,"value":[0.36998605196364226,1.79,0.03,0.029998450218699874,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.008102,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.008102,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.008102,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.008102,"value":[8.0,64.5455,10.0,5.0,0.0,0.9901,95.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.014536,"value":[0.0,0.0,0.02,0.029998450218699874,0.02,0.08999845021840884,0.03999845021840884,0.08999380087392637],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.014536,"value":[0.02,0.0,1.53,0.14999845021869987,0.65,0.3,2.92,0.3299752034957055],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.014536,"value":[0.5699860519636423,3.2,0.07,0.039998450218699876,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.014536,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.014536,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.014536,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.014536,"value":[11.8812,63.1579,4.0,4.0,35.6436,29.0,100.0,20.4082],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.020766,"value":[0.01,0.0,0.03,0.04999845021869987,0.02,0.09999845021840884,0.03999845021840884,0.08999380087392637],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.020766,"value":[0.02,0.0,3.44,0.18999845021869988,0.99,0.7,2.96,0.3299752034957055],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.020766,"value":[0.7699860519636422,4.62,0.09,0.04999845021869987,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.020766,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.020766,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.020766,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.020766,"value":[8.0,63.3929,100.0,4.9505,11.0,28.4314,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.026704,"value":[0.01,0.0,0.05,0.05999845021869987,0.04,0.09999845021840884,0.03999845021840884,0.08999380087392637],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.026704,"value":[0.03,0.0,5.37,0.22999845021869988,1.01,0.71,2.96,0.3299752034957055],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.026704,"value":[0.9699860519636423,6.02,0.1,0.05999845021869987,0.01,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.026704,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.026704,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.026704,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.026704,"value":[11.0,59.6639,100.0,3.9604,1.9802,0.9901,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":11.485422,"value":[0.014545202843728476,0.004545202843728475,0.05,0.06999845021869987,0.07,0.10999845021840883,0.03999845021840884,0.11999380087392637],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":11.485422,"value":[0.03,0.004545202843728475,6.53089851943776,0.2781792615936138,1.03,0.71,2.96,0.4899752034957055],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":11.485422,"value":[1.1163476747181267,6.850901288156383,0.11454520284376485,0.05999845021869987,0.01,0.004545202843764855,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":11.485422,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":11.485422,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":11.485422,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":11.485422,"value":[9.0,31.1927,43.4343,4.0,0.0,0.9901,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.002449,"value":[0.0,0.01,0.019996900436817667,0.009998450218699873,0.0,0.009998450218408833,0.009998450218408833,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.002449,"value":[0.009998450218408833,0.01,0.009998450218699873,0.039998450218699876,0.26,0.01,0.009998450218408833,0.15],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.002449,"value":[0.20998140262439846,0.07,0.01,0.009998450218699873,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.002449,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.002449,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.002449,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.002449,"value":[9.1837,9.8039,1.0,3.0612,26.5306,0.9901,0.0,17.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.008102,"value":[0.01,0.01,0.019996900436817667,0.01999845021869987,0.01,0.009998450218408833,0.009998450218408833,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.008102,"value":[0.019998450218408834,0.01,0.009998450218699873,0.09999845021869987,1.24,0.01,0.009998450218408833,0.16],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.008102,"value":[0.38998140262439845,0.31,0.01,0.01999845021869987,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.008102,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.008102,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.008102,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.008102,"value":[10.7843,12.1212,0.0,4.0,50.5051,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.014536,"value":[0.01,0.01,0.02999690043681767,0.05999845021869987,0.01,0.029998450218408836,0.009998450218408833,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.014536,"value":[0.029998450218408836,0.02,0.009998450218699873,0.13999845021869986,1.32,0.5,0.5299984502184089,0.17],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.014536,"value":[0.5999814026243985,0.58,0.01,0.029998450218699874,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.014536,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.014536,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.014536,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.014536,"value":[13.1313,13.0,1.0,4.0,0.0,7.0707,52.4752,0.9901],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.020766,"value":[0.01,0.01,0.03999690043681767,0.07999845021869988,0.01,0.03999845021840884,0.029998450218408836,0.07],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.020766,"value":[0.03999845021840884,0.02,0.01999845021869987,0.18999845021869988,1.32,0.68,1.5499984502184088,0.38],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.020766,"value":[0.8899814026243985,0.81,0.01,0.039998450218699876,0.0,0.0,0.0,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.020766,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.020766,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.020766,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.020766,"value":[14.0,13.0,0.0,5.0,0.0,17.1717,45.4545,6.9307],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.026704,"value":[0.02,0.01,0.03999690043681767,0.09999845021869987,0.01,0.059998450218408834,0.029998450218408836,0.07],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.026704,"value":[0.04999845021840883,0.02,0.01999845021869987,0.24999845021869987,1.32,1.93,1.5499984502184088,0.42],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.026704,"value":[1.1199814026243984,1.09,0.01,0.04999845021869987,0.0,0.0,0.0,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.026704,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.026704,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.026704,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.026704,"value":[13.7255,16.1616,0.0,3.0303,0.0,64.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":11.485422,"value":[0.02,0.014545202843728476,0.04999690043681767,0.11454365306242835,0.01,0.06454365306213732,0.029998450218408836,0.07909040568745695],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":11.485422,"value":[0.04999845021840883,0.024545202843728475,0.3863545355311362,0.2936340587493032,1.32,2.3445452028437286,1.5545436530621373,0.4563616227501188],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":11.485422,"value":[1.299069039588794,1.238180811375023,0.024545202843728475,0.05999845021869987,0.0,0.0045452028437557605,0.0,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":11.485422,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":11.485422,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":11.485422,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":11.485422,"value":[13.2653,5.9406,30.6931,5.0,0.9804,2.9703,1.0,10.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"} + ] + }, + { + "group": "selftests-drivers-net-perf", + "test": "devmem-py", + "subtest": "devmem-kperf", + "link": "/artifacts/1/test-outputs/0-devmem-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 328.177 + }, + "regression_value": 328.177, + "value_count": 1, + "regression_result": { + "status": "ok", + "current": 328.177, + "baseline_samples": 1, + "baseline": 215.378, + "threshold": 21.5378, + "delta_percent": 52.372573 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.189528 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 2.741172 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 2.463136 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 333.128203 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 5.351488 + }, + "value_count": 17 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.029922 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 1.226591 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 1.426115 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 2.837191 + }, + "value_count": 17 + } + ], + "attempt": 0, + "result": "pass", + "id": 3, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-47", + "branch_date": "2026-09-25--21-47", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:47:16.227591+00:00", + "full_json": [ + {"goal":"higher","name":"throughput","value":328.17699999999996,"origin":"selftest"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.002154,"value":[0.0,0.0,0.009999477544915862,0.0,0.01,0.01,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.002154,"value":[0.009999477544915862,0.0,0.0,0.0,0.029999477544333786,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.002154,"value":[0.18999425299465655,0.0,0.0,0.0,0.019999477544915864,0.0,0.01,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.002154,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.002154,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.002154,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.002154,"value":[8.0808,0.0,0.0,0.0,3.9604,0.9901,0.9901,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.00713,"value":[0.0,0.0,0.019999477544915864,0.01,0.36,0.01,0.0,0.01],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.00713,"value":[0.019999477544915864,0.01,0.01,0.01,1.199999477544334,0.0,0.03,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.00713,"value":[0.32999425299465657,0.0,0.0,0.0,0.019999477544915864,0.0,0.01,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.00713,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.00713,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.00713,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.00713,"value":[8.0808,0.0,1.9802,0.0,87.0,0.0,2.0202,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.013992,"value":[0.0,0.0,0.029999477544915863,0.01,0.84,0.01,0.0,0.01],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.013992,"value":[0.029999477544915863,0.01,0.01,0.02,2.699999477544334,0.0,0.04,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.013992,"value":[0.4799942529946566,0.0,0.0,0.0,0.029999477544915863,0.0,0.01,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.013992,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.013992,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.013992,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.013992,"value":[8.0808,0.0,0.9901,0.0,100.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.01841,"value":[0.0,0.0,0.029999477544915863,0.01,1.08,0.01,0.0,0.01],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.01841,"value":[0.029999477544915863,0.02,0.02,0.02,3.4799994775443337,0.0,0.05,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.01841,"value":[0.6399942529946565,0.0,0.0,0.0,0.03999947754491586,0.03,0.01,0.04],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.01841,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.01841,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.01841,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.01841,"value":[7.0,0.0,0.9901,0.0,3.0303,0.0,1.9802,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.023637,"value":[0.01,0.01,0.029999477544915863,0.02,1.08,0.01,0.0,0.01],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.023637,"value":[0.03999947754491586,0.02,0.02,0.06,3.4799994775443337,0.0,0.05,0.01],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.023637,"value":[0.7599942529946566,0.01,0.01,0.0,0.04999947754491586,0.03,0.01,0.04],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.023637,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.023637,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.023637,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.023637,"value":[6.0606,0.0,1.0,4.9505,1.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":12.028791,"value":[0.01,0.04,0.03999947754491586,0.02,1.09,0.01,0.0,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":12.028791,"value":[0.04999947754491586,0.03,0.03,0.07,3.529999477544334,0.0,0.06,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":12.028791,"value":[0.9199942529946565,0.01,0.01,0.01,0.04999947754491586,0.03,0.02,0.04],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":12.028791,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":12.028791,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":12.028791,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":12.028791,"value":[10.0,0.0,1.0,1.9802,6.8627,0.0,1.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":14.03477,"value":[0.01,0.04,0.03999947754491586,0.03,1.21,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":14.03477,"value":[0.04999947754491586,0.03,0.03,0.08,3.8599994775443336,0.0,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":14.03477,"value":[1.1499942529946565,0.05,0.05,0.05,0.059999477544915865,0.07,0.05,0.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":14.03477,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":14.03477,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":14.03477,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":14.03477,"value":[9.0909,0.0,0.0,0.9901,4.0,0.0,1.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":16.040004,"value":[0.01,0.06,0.04999947754491586,0.04,1.24,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":16.040004,"value":[0.059999477544915865,0.73,1.0,1.04,3.909999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":16.040004,"value":[1.3799942529946565,0.06,0.05,0.05,0.06999947754491587,0.73,0.76,1.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":16.040004,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":16.040004,"value":668.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":16.040004,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":16.040004,"value":[16.0,77.3196,98.0,98.9796,6.8627,68.0412,71.875,99.0099],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":18.045577,"value":[0.01,0.1,0.08999947754491586,0.08,1.25,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":18.045577,"value":[0.059999477544915865,2.12,2.95,2.98,3.9599994775443337,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":18.045577,"value":[1.6699942529946565,0.08,0.07,0.07,0.07999947754491586,2.05,2.02,3.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":18.045577,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":18.045577,"value":668.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":18.045577,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":18.045577,"value":[15.0,78.7234,100.0,100.0,2.9703,69.7917,64.5833,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":20.051262,"value":[0.01,0.12,0.10999947754491586,0.11,1.26,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":20.051262,"value":[0.06999947754491587,3.58,4.91,4.94,4.019999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":20.051262,"value":[1.9299942529946565,0.11,0.08,0.09,0.08999947754491586,3.43,3.29,5.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":20.051262,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":20.051262,"value":668.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":20.051262,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":20.051262,"value":[13.0,79.3814,100.0,100.0,5.8824,71.5789,65.3061,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":22.056729,"value":[0.01,0.16,0.13999947754491587,0.14,1.27,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":22.056729,"value":[0.06999947754491587,4.96,6.83,6.86,4.069999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":22.056729,"value":[2.2099942529946563,0.13,0.1,0.1,0.09999947754491587,4.67,4.47,7.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":22.056729,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":22.056729,"value":668.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":22.056729,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":22.056729,"value":[15.8416,77.8947,99.0,99.0099,4.902,61.2245,58.5859,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":24.062288,"value":[0.02,0.18,0.16999947754491587,0.18,1.28,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":24.062288,"value":[0.07999947754491586,6.47,8.77,8.78,4.109999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":24.062288,"value":[2.4799942529946564,0.16,0.11,0.12,0.09999947754491587,5.89,5.65,9.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":24.062288,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":24.062288,"value":680.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":24.062288,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":24.062288,"value":[15.0,86.4583,100.0,100.0,3.0,63.6364,59.4059,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":26.067737,"value":[0.02,0.21,0.19999947754491587,0.21,1.29,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":26.067737,"value":[0.07999947754491586,8.05,10.7,10.72,4.159999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":26.067737,"value":[2.7499942529946564,0.18,0.13,0.13,0.10999947754491586,7.12,6.82,11.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":26.067737,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":26.067737,"value":680.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":26.067737,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":26.067737,"value":[15.0,85.567,99.0099,99.0099,4.0404,62.0,59.0,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":28.074178,"value":[0.02,0.24,0.21999947754491586,0.24,1.3,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":28.074178,"value":[0.07999947754491586,9.66,12.64,12.63,4.199999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":28.074178,"value":[3.0199942529946564,0.21,0.14,0.15,0.11999947754491586,8.35,7.99,13.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":28.074178,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":28.074178,"value":680.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":28.074178,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":28.074178,"value":[14.1414,85.0,99.0,99.0,2.0408,61.3861,58.4158,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":30.080062,"value":[0.02,0.27,0.24999947754491586,0.26,1.31,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":30.080062,"value":[0.08999947754491586,11.25,14.55,14.54,4.249999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":30.080062,"value":[3.2799942529946566,0.23,0.16,0.17,0.12999947754491586,9.56,9.12,15.08],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":30.080062,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":30.080062,"value":680.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":30.080062,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":30.080062,"value":[14.1414,85.0,97.0297,97.0297,2.0,61.8557,57.1429,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":32.085669,"value":[0.02,0.27,0.24999947754491586,0.26,1.61,0.01,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":32.085669,"value":[0.08999947754491586,11.26,14.68,14.54,5.159999477544334,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":32.085669,"value":[3.4499942529946566,0.23,0.16,0.17,0.13999947754491587,9.56,9.12,15.09],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":32.085669,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":32.085669,"value":680.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":32.085669,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":32.085669,"value":[8.0,0.0,0.9901,0.0,48.0392,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":33.986001,"value":[0.028948115508828778,0.27,0.24999947754491586,0.26,1.6189481155085377,0.01894811550868326,0.01,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":33.986001,"value":[0.08999947754491586,11.27,14.68,14.54,5.205791939578485,0.01,0.07,0.02],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":33.986001,"value":[3.6173716390971093,0.2578962310174393,0.1778962310174029,0.18789623101743927,0.13999947754491587,9.577896231017403,9.137896231017402,15.107896231017403],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":33.986001,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":33.986001,"value":680.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":33.986001,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":33.986001,"value":[12.8713,1.9608,2.0,2.0,5.0505,2.9412,1.9802,2.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.002154,"value":[0.0,0.009999477544915862,0.009999477544915862,0.0,0.019999477544915864,0.01,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.002154,"value":[0.0,0.0,0.009999477544915862,0.0,0.03,0.009999477544915862,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.002154,"value":[0.20999477544799447,0.01,0.0,0.0,0.009999477544952243,0.009999477544915862,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.002154,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.002154,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.002154,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.002154,"value":[11.2245,1.0,0.0,0.0,4.0,1.0,2.9412,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.00713,"value":[0.0,0.019999477544915864,0.009999477544915862,0.0,0.029999477544915863,0.01,0.02,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.00713,"value":[0.01,0.01,0.009999477544915862,0.01,0.03,0.009999477544915862,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.00713,"value":[0.35999477544799446,0.01,0.0,0.01,0.009999477544952243,0.009999477544915862,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.00713,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.00713,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.00713,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.00713,"value":[8.0,1.0,0.0,1.9608,0.0,0.0,1.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.013992,"value":[0.01,0.029999477544915863,0.009999477544915862,0.01,0.029999477544915863,0.01,0.02,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.013992,"value":[0.02,0.01,0.009999477544915862,0.01,0.04,0.009999477544915862,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.013992,"value":[0.4999947754479945,0.01,0.0,0.01,0.009999477544952243,0.009999477544915862,0.01,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.013992,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.013992,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.013992,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.013992,"value":[9.0,2.0,0.0,1.0,1.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.01841,"value":[0.01,0.029999477544915863,0.019999477544915864,0.01,0.03999947754491586,0.01,0.02,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.01841,"value":[0.03,0.02,0.019999477544915864,0.01,0.04,0.009999477544915862,0.02,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.01841,"value":[0.6699947754479945,0.01,0.01,0.01,0.009999477544952243,0.019999477544915864,0.02,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.01841,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.01841,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.01841,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.01841,"value":[9.0,0.0,1.0,0.0,1.0,0.9901,0.0,1.9608],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.023637,"value":[0.03,0.03999947754491586,0.019999477544915864,0.01,0.03999947754491586,0.02,0.03,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.023637,"value":[0.05,0.02,0.019999477544915864,0.02,0.04,0.029999477544915863,0.02,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.023637,"value":[0.7999947754479945,0.01,0.01,0.02,0.009999477544952243,0.019999477544915864,0.02,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.023637,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.023637,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.023637,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.023637,"value":[9.0909,0.0,0.0,1.0,0.0,2.0,1.0,0.9901],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":12.028791,"value":[0.03,0.04999947754491586,0.029999477544915863,0.02,0.04999947754491586,0.03,0.03,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":12.028791,"value":[0.05,0.02,0.019999477544915864,0.02,0.05,0.03999947754491586,0.03,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":12.028791,"value":[0.9799947754479945,0.01,0.01,0.02,0.009999477544952243,0.019999477544915864,0.03,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":12.028791,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":12.028791,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":12.028791,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":12.028791,"value":[11.1111,1.0,1.0,0.0,0.9901,1.0,1.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":14.03477,"value":[0.03,0.059999477544915865,0.029999477544915863,0.02,0.04999947754491586,0.03,0.03,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":14.03477,"value":[0.06,0.05,0.029999477544915863,0.02,0.05,0.04999947754491586,0.03,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":14.03477,"value":[1.2199947754479945,0.01,0.02,0.02,0.009999477544952243,0.019999477544915864,0.03,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":14.03477,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":14.03477,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":14.03477,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":14.03477,"value":[13.7255,1.9802,0.9901,0.0,0.9901,0.9901,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":16.040004,"value":[0.03,0.059999477544915865,0.029999477544915863,0.02,0.04999947754491586,0.03,0.04,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":16.040004,"value":[0.06,0.05,0.20999947754491585,0.27,0.05,0.29999947754491585,0.03,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":16.040004,"value":[1.5899947754479944,0.19,0.03,0.32,0.17999947754495224,0.30999947754491586,0.19,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":16.040004,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":16.040004,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":16.040004,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":16.040004,"value":[34.0,20.202,19.1919,60.2151,17.0,60.8696,17.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":18.045577,"value":[0.03,0.059999477544915865,0.029999477544915863,0.02,0.04999947754491586,0.03,0.04,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":18.045577,"value":[0.06,0.05,0.5999994775449159,0.71,0.05,0.7699994775449158,0.04,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":18.045577,"value":[2.2499947754479943,0.55,0.04,0.94,0.5099994775449522,0.9399994775449159,0.52,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":18.045577,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":18.045577,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":18.045577,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":18.045577,"value":[35.0,19.0,21.7822,58.6957,17.0,59.5745,17.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":20.051262,"value":[0.04,0.059999477544915865,0.029999477544915863,0.02,0.059999477544915865,0.03,0.04,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":20.051262,"value":[0.07,0.05,0.9699994775449159,1.16,0.05,1.1799994775449159,0.04,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":20.051262,"value":[2.8499947754479944,0.89,0.05,1.54,0.8399994775449522,1.5799994775449158,0.84,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":20.051262,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":20.051262,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":20.051262,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":20.051262,"value":[31.6327,18.0,19.0,58.5106,18.6275,58.9474,16.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":22.056729,"value":[0.04,0.059999477544915865,0.029999477544915863,0.02,0.059999477544915865,0.03,0.04,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":22.056729,"value":[0.08,0.05,1.509999477544916,1.85,0.05,1.8499994775449158,0.05,0.62],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":22.056729,"value":[3.5599947754479944,1.27,0.07,2.1,1.1399994775449522,2.139999477544916,1.13,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":22.056729,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":22.056729,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":22.056729,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":22.056729,"value":[38.7755,20.0,30.6931,70.8333,14.0,70.2128,15.0,34.6154],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":24.062288,"value":[0.04,0.059999477544915865,0.029999477544915863,0.02,0.059999477544915865,0.03,0.05,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":24.062288,"value":[0.09,0.05,2.109999477544916,2.48,0.05,2.499999477544916,0.05,1.19],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":24.062288,"value":[4.209994775447995,1.66,0.09,2.68,1.4599994775449523,2.709999477544916,1.43,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":24.062288,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":24.062288,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":24.062288,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":24.062288,"value":[34.0,21.2121,32.6733,66.6667,16.0,65.9574,15.0,24.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":26.067737,"value":[0.04,0.059999477544915865,0.029999477544915863,0.02,0.059999477544915865,0.03,0.05,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":26.067737,"value":[0.1,0.05,2.7399994775449157,3.1,0.06,3.129999477544916,0.05,1.69],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":26.067737,"value":[4.869994775447995,2.05,0.12,3.27,1.7699994775449523,3.289999477544916,1.73,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":26.067737,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":26.067737,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":26.067737,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":26.067737,"value":[37.0,21.0,34.3137,68.0851,15.0,67.7083,14.1414,26.4706],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":28.074178,"value":[0.04,0.06999947754491587,0.03999947754491586,0.02,0.06999947754491587,0.03,0.05,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":28.074178,"value":[0.1,0.06,3.429999477544916,3.83,0.06,3.869999477544916,0.05,1.88],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":28.074178,"value":[5.579994775447995,2.45,0.15,3.82,2.069999477544952,3.829999477544916,2.02,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":28.074178,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":28.074178,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":28.074178,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":28.074178,"value":[37.1134,21.7822,38.4615,69.4737,16.0,71.875,14.8515,0.9901],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":30.080062,"value":[0.04,0.06999947754491587,0.03999947754491586,0.02,0.06999947754491587,0.03,0.05,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":30.080062,"value":[0.11,0.06,4.029999477544916,4.47,0.06,4.529999477544916,0.05,2.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":30.080062,"value":[6.2199947754479945,2.85,0.17,4.4,2.3799994775449522,4.389999477544916,2.31,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":30.080062,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":30.080062,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":30.080062,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":30.080062,"value":[34.3434,21.7822,30.6931,65.5914,16.0,67.0213,14.0,18.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":32.085669,"value":[0.05,0.06999947754491587,0.03999947754491586,0.05,0.06999947754491587,0.04,0.06,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":32.085669,"value":[0.12,0.06,4.029999477544916,4.49,0.07,4.5399994775449155,0.05,2.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":32.085669,"value":[6.459994775447995,2.85,0.17,4.4,2.3799994775449522,4.389999477544916,2.31,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":32.085669,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":32.085669,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":32.085669,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":32.085669,"value":[14.8515,0.0,0.0,4.9505,0.0,1.9802,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":33.986001,"value":[0.05,0.06999947754491587,0.03999947754491586,0.05,0.06999947754491587,0.04,0.07,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":33.986001,"value":[0.1289481155088288,0.07,4.049999477544916,4.49,0.07,4.5399994775449155,0.06,2.06],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":33.986001,"value":[6.640527815027163,2.85,0.18,4.41,2.3799994775449522,4.389999477544916,2.31,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":33.986001,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":33.986001,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":33.986001,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":33.986001,"value":[10.0,0.0,0.0,0.0,0.0,0.0,0.0,0.9901],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"} + ] + }, + { + "group": "selftests-drivers-net-perf", + "test": "iperf-py", + "subtest": "iperf-tcp", + "link": "/artifacts/1/test-outputs/1-iperf-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 33.511069 + }, + "regression_value": 33.511069, + "value_count": 1, + "regression_result": { + "status": "regression", + "current": 33.511069, + "baseline_samples": 1, + "baseline": 57.294036, + "threshold": 5.729404, + "delta_percent": -41.510371 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.109655 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 1.116483 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.852315 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 2.064525 + }, + "value_count": 6 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.159968 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.668028 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.239293 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 0.932169 + }, + "value_count": 6 + } + ], + "attempt": 0, + "result": "pass", + "id": 4, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-47", + "branch_date": "2026-09-25--21-47", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:47:16.227591+00:00", + "full_json": [ + {"goal":"higher","name":"throughput","value":33.51106915715046,"origin":"selftest"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.00039,"value":[0.0,0.0,0.009999564483587164,0.009999564483878202,0.01,0.0,0.0,0.009999564483732683],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.00039,"value":[0.0,0.0,0.009999564483587164,0.059998693452216686,0.59,0.0,0.0,0.009999564483587164],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.00039,"value":[0.159996515866369,0.42,0.0,0.01,0.0,0.0,0.02,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.00039,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.00039,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.00039,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.00039,"value":[9.0909,42.0,0.0,4.9505,60.0,0.0,3.9604,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.006295,"value":[0.01,0.0,0.009999564483587164,0.019999564483878202,0.02,0.0,0.0,0.009999564483732683],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.006295,"value":[0.01,0.0,0.009999564483587164,0.10999869345221669,2.57,0.0,0.0,0.009999564483587164],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.006295,"value":[0.359996515866369,1.8,0.0,0.02,0.01,0.0,0.09,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.006295,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.006295,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.006295,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.006295,"value":[11.0,68.6869,0.0,4.0,100.0,0.0,5.9406,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.012399,"value":[0.01,0.0,0.009999564483587164,0.0399995644838782,0.02,0.0,0.0,0.009999564483732683],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.012399,"value":[0.02,0.0,0.009999564483587164,0.1499986934522167,3.84,0.0,0.0,0.7199995644835872],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.012399,"value":[0.589996515866369,3.17,0.0,0.03,0.01,0.0,0.16,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.012399,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.012399,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.012399,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.012399,"value":[12.1212,69.0,0.9901,3.9604,27.2727,0.0,5.9406,72.449],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.018447,"value":[0.01,0.0,0.009999564483587164,0.049999564483878205,0.03,0.0,0.0,0.019999564483732683],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.018447,"value":[0.02,0.0,0.009999564483587164,0.19999869345221669,3.85,0.0,0.0,2.719999564483587],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.018447,"value":[0.789996515866369,4.57,0.0,0.04,0.01,0.0,0.23,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.018447,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.018447,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.018447,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.018447,"value":[10.8911,70.297,0.0,5.0505,0.9901,0.0,6.9307,100.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.024748,"value":[0.01,0.0,0.019999564483587164,0.0599995644838782,0.04,0.19,0.0,0.019999564483732683],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.024748,"value":[0.02,0.0,0.019999564483587164,0.24999869345221667,3.85,0.27,0.01,4.619999564483587],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.024748,"value":[1.009996515866369,5.98,0.0,0.05,0.01,0.0,0.3,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.024748,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.024748,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.024748,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.024748,"value":[12.1212,72.0,0.0,3.0,0.0,23.2323,6.0,90.099],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":11.430368,"value":[0.01,0.0,0.0500750953584793,0.06401467065908946,0.07,0.19,0.0,0.019999564483732683],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":11.430368,"value":[0.05409063704981236,0.0,1.0366642361832783,0.27401379962684586,3.87,0.28,0.01,4.619999564483587],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":11.430368,"value":[1.18613247144036,6.646287017324867,0.004015106174992979,0.06,0.01,0.0,0.3,0.0],"host":"local","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":11.430368,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":11.430368,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":11.430368,"value":0.0,"host":"local","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":11.430368,"value":[16.1616,17.4312,50.0,2.0202,0.0,0.0,0.0,0.0],"host":"local","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":0.0,"value":[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":0.0,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.time.usr","type":"per-cpu","time":2.00039,"value":[0.009999564483587164,0.0,0.0,0.26999956448358714,0.0,0.01,0.019999128967465366,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":2.00039,"value":[0.009999564483587164,0.0,0.009999564483587164,0.09999956448358717,0.01,0.05,0.03999869345105253,0.25],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":2.00039,"value":[0.22999564483761786,0.09,0.0,0.01,0.0,0.0,0.0,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":2.00039,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":2.00039,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":2.00039,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":2.00039,"value":[13.1313,10.0,0.0,36.6337,0.9901,6.0,0.9901,25.4717],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":4.006295,"value":[0.009999564483587164,0.0,0.0,0.27999956448358715,0.0,0.03,0.019999128967465366,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":4.006295,"value":[0.009999564483587164,0.0,0.009999564483587164,0.09999956448358717,0.48,0.16,0.39999869345105254,0.65],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":4.006295,"value":[0.41999564483761787,0.38,0.0,0.01,0.0,0.0,0.0,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":4.006295,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":4.006295,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":4.006295,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":4.006295,"value":[9.901,15.0,0.0,0.0,18.1818,0.0,24.0385,14.5631],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":6.012399,"value":[0.009999564483587164,0.0,0.0,0.28999956448358716,0.01,0.03,0.019999128967465366,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":6.012399,"value":[0.009999564483587164,0.0,0.14999956448358717,0.09999956448358717,0.93,0.16,0.4299986934510525,1.16],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":6.012399,"value":[0.5999956448376179,0.65,0.01,0.01,0.0,0.0,0.01,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":6.012399,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":6.012399,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":6.012399,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":6.012399,"value":[9.1837,13.1313,14.8515,1.0,26.0,0.0,0.0,20.202],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":8.018447,"value":[0.019999564483587164,0.0,0.01,0.29999956448358717,0.01,0.06,0.029999128967465368,0.02],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":8.018447,"value":[0.019999564483587164,0.0,0.8699995644835872,0.10999956448358716,0.93,0.33,0.4299986934510525,1.16],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":8.018447,"value":[0.7799956448376179,0.84,0.02,0.01,0.0,0.0,0.01,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":8.018447,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":8.018447,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":8.018447,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":8.018447,"value":[9.0909,9.0909,38.0,0.9901,0.0,12.1212,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":10.024748,"value":[0.019999564483587164,0.02,0.01,0.29999956448358717,0.01,0.06,0.029999128967465368,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":10.024748,"value":[0.039999564483587165,0.0,1.5699995644835871,0.10999956448358716,0.93,0.33,0.4299986934510525,1.16],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":10.024748,"value":[0.9199956448376179,1.04,0.04,0.01,0.0,0.0,0.01,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":10.024748,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":10.024748,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":10.024748,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":10.024748,"value":[9.901,10.8911,37.5,0.0,0.0,0.0,0.0,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"}, + {"name":"cpu.time.usr","type":"per-cpu","time":11.430368,"value":[0.029999564483587163,0.02401510617492022,0.01401510617492022,0.3040146706585074,0.02606042469997192,0.06401510617492022,0.029999128967465368,0.03],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sys","type":"per-cpu","time":11.430368,"value":[0.039999564483587165,0.012045318525051698,2.0803016879834466,0.11401467065850739,0.9380302123498404,0.3681057432250236,0.43401379962597275,1.16],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"cpu.time.sirq","type":"per-cpu","time":11.430368,"value":[1.0802222374640404,1.126060424699972,0.05401510617492022,0.02,0.0,0.0,0.01,0.01],"host":"remote","origin":"system-monitor","summarize":"p90","transform":"rate","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_scale":100,"display_label":"Percent of one CPU"}, + {"name":"nic.rx.dropped","type":"scalar","time":11.430368,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.rx.missed_errors","type":"scalar","time":11.430368,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"nic.tx.dropped","type":"scalar","time":11.430368,"value":0.0,"host":"remote","origin":"system-monitor","summarize":"max","transform":"rate","display_range":{"min":0}}, + {"name":"cpu.utilization","type":"per-cpu","time":11.430368,"value":[15.1515,8.9109,21.7822,2.9412,6.0,8.0,0.9901,0.0],"host":"remote","origin":"system-monitor","summarize":"p90","aggregation":"busy-core-equivalents","display_range":{"min":0,"max":100},"display_label":"Percent of one CPU"} + ] + } +] diff --git a/ui/mock/result-metrics.json b/ui/mock/result-metrics.json new file mode 100644 index 00000000..4ae551a1 --- /dev/null +++ b/ui/mock/result-metrics.json @@ -0,0 +1,1184 @@ +[ + { + "group": "selftests-drivers-net-perf", + "test": "devmem-py", + "subtest": "devmem-kperf", + "link": "/artifacts/0/test-outputs/0-devmem-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 215.378 + }, + "regression_value": 215.378, + "value_count": 1, + "regression_result": { + "status": "insufficient-data", + "current": 215.378, + "baseline_samples": 0 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.079788 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 3.050889 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 2.148583 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 633.769492 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 5.486676 + }, + "value_count": 17 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.054877 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.573283 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.762722 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 1.29674 + }, + "value_count": 17 + } + ], + "attempt": 0, + "result": "pass", + "id": 1, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-33", + "branch_date": "2026-09-25--21-33", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:33:21.292718+00:00" + }, + { + "group": "selftests-drivers-net-perf", + "test": "iperf-py", + "subtest": "iperf-tcp", + "link": "/artifacts/0/test-outputs/1-iperf-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 57.294036 + }, + "regression_value": 57.294036, + "value_count": 1, + "regression_result": { + "status": "insufficient-data", + "current": 57.294036, + "baseline_samples": 0 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.074759 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 1.554998 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.827338 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 2.680909 + }, + "value_count": 6 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.049845 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.737702 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.269162 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 1.015569 + }, + "value_count": 6 + } + ], + "attempt": 0, + "result": "pass", + "id": 2, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-33", + "branch_date": "2026-09-25--21-33", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:33:21.292718+00:00" + }, + { + "group": "selftests-drivers-net-perf", + "test": "devmem-py", + "subtest": "devmem-kperf", + "link": "/artifacts/1/test-outputs/0-devmem-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 328.177 + }, + "regression_value": 328.177, + "value_count": 1, + "regression_result": { + "status": "ok", + "current": 328.177, + "baseline_samples": 1, + "baseline": 215.378, + "threshold": 21.5378, + "delta_percent": 52.372573 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.189528 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 2.741172 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 2.463136 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 333.128203 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 5.351488 + }, + "value_count": 17 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 0.029922 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 1.226591 + }, + "value_count": 17 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 18, + "cpus": 8, + "summary": { + "p90": 1.426115 + }, + "value_count": 17 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 18, + "summary": { + "max": 0.0 + }, + "value_count": 17 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 17, + "cpus": 8, + "summary": { + "p90": 2.837191 + }, + "value_count": 17 + } + ], + "attempt": 0, + "result": "pass", + "id": 3, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-47", + "branch_date": "2026-09-25--21-47", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:47:16.227591+00:00" + }, + { + "group": "selftests-drivers-net-perf", + "test": "iperf-py", + "subtest": "iperf-tcp", + "link": "/artifacts/1/test-outputs/1-iperf-py/", + "short_json": [ + { + "goal": "higher", + "name": "throughput", + "origin": "selftest", + "samples": 1, + "summarize": "last", + "display_label": "Gbps", + "summary": { + "last": 33.511069 + }, + "regression_value": 33.511069, + "value_count": 1, + "regression_result": { + "status": "regression", + "current": 33.511069, + "baseline_samples": 1, + "baseline": 57.294036, + "threshold": 5.729404, + "delta_percent": -41.510371 + } + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.109655 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 1.116483 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.852315 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "local", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "local", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 2.064525 + }, + "value_count": 6 + }, + { + "name": "cpu.time.usr", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.159968 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sys", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.668028 + }, + "value_count": 6 + }, + { + "name": "cpu.time.sirq", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "transform": "rate", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_scale": 100, + "display_label": "Percent of one CPU", + "samples": 7, + "cpus": 8, + "summary": { + "p90": 0.239293 + }, + "value_count": 6 + }, + { + "name": "nic.rx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.rx.missed_errors", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "nic.tx.dropped", + "type": "scalar", + "host": "remote", + "origin": "system-monitor", + "summarize": "max", + "transform": "rate", + "display_range": { + "min": 0 + }, + "samples": 7, + "summary": { + "max": 0.0 + }, + "value_count": 6 + }, + { + "name": "cpu.utilization", + "type": "per-cpu", + "host": "remote", + "origin": "system-monitor", + "summarize": "p90", + "aggregation": "busy-core-equivalents", + "display_range": { + "min": 0, + "max": 100 + }, + "display_label": "Percent of one CPU", + "samples": 6, + "cpus": 8, + "summary": { + "p90": 0.932169 + }, + "value_count": 6 + } + ], + "attempt": 0, + "result": "pass", + "id": 4, + "branch": "ksft-perf-rfc-v2/net-next-2026-09-25--21-47", + "branch_date": "2026-09-25--21-47", + "remote": "local", + "executor": "root@dut.example.com", + "start": "2026-09-25T21:47:16.227591+00:00" + } +] From 8698586d3f7b815ef79b0b45a8c535858b17e295 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:21:18 -0700 Subject: [PATCH 09/18] ui: add performance history charts Plot compact metric summaries across test runs. Group series by producer metadata and link each point to its complete observations. Query the history of the selected test case only, and label series with their origin when a chart has metrics of more than one origin. Load a data file named in the URL, like the fixtures, only from this site. Add query/metrics to the offline assets of scripts/ui_assets.sh. Signed-off-by: Stanislav Fomichev --- scripts/ui_assets.sh | 1 + ui/favicon-perf.png | Bin 0 -> 477 bytes ui/metrics.html | 41 ++++++ ui/metrics.js | 327 +++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 369 insertions(+) create mode 100644 ui/favicon-perf.png create mode 100644 ui/metrics.html create mode 100644 ui/metrics.js diff --git a/scripts/ui_assets.sh b/scripts/ui_assets.sh index e2d192f5..725d3702 100755 --- a/scripts/ui_assets.sh +++ b/scripts/ui_assets.sh @@ -31,6 +31,7 @@ ASSETS=( "query/stability?auto=1" "query/remotes" "query/device-info" + "query/metrics" "mc/get_machine_info?caller=status-ui" ) diff --git a/ui/favicon-perf.png b/ui/favicon-perf.png new file mode 100644 index 0000000000000000000000000000000000000000..cb5cb1397485736307bd183d8efc6587c6a181f8 GIT binary patch literal 477 zcmV<30V4j1P)BY{7)^~2HvEUgD^zdu&izrSrhgZb<4 z1Hf;!VPM{mZJ;fXcA)J6v~5WTYJ-Q(S`xgw0`M4!Y}7}=h4#aRJ{T)F{e_PI=U^~2 zcsC@t$8y2o5wBO9uRIuc!<{n-gQuDRAln5efM6I8hSk9^PeRYfsBE~k4HE!C!I6Ys z+#7zjeYy^!VAw1ewod@i&#*5#Swts!bW)Je7lR6=2LR5%mV)=>1p#2_HafW{p%;US zgaYfPVxClIOK`OYqLcIJIaqLV5**(Lz!)iVLQXQ0$p zmD;wYwvmKhB><|aSepQ#R|$X$hP4R*dPM+>3p22ICP4QROaOEX@(BQXF;WrYIp7lj ze|8Q`029CjFaf**0KNrl3ijmrNEZ^kg4Go4$@5Vkr}z*h2tne<$7%07;eUL;o$gl; T + + + + NIPA result metrics + + + + + + + + + + + +
+

Performance history

+ +

Loading...

+
+ + diff --git a/ui/metrics.js b/ui/metrics.js new file mode 100644 index 00000000..95daee03 --- /dev/null +++ b/ui/metrics.js @@ -0,0 +1,327 @@ +let metric_runs = []; +let metric_data_source = null; + +const chart_colors = [ + '#3366cc', '#dc3912', '#109618', '#ff9900', '#990099', + '#0099c6', '#dd4477', '#66aa00', '#b82e2e', '#316395' +]; +// URL parameters selecting the test case and runner of the history +const metric_case_keys = ['remote', 'executor', 'group', 'test', 'subtest']; + +function runner_name(row) +{ + return `${row.remote}:${row.executor}`; +} + +function case_name(row) +{ + return `${row.group}:${row.test}:${row.subtest}`; +} + +function selected_case(params) +{ + const selected = {}; + for (const key of metric_case_keys) { + if (!params.get(key)) + return null; + selected[key] = params.get(key); + } + return selected; +} + +function filtered_runs() +{ + const params = new URLSearchParams(window.location.search); + const selected_branch = params.get('branch'); + const selected = selected_case(params); + + if (!selected) + return []; + + return metric_runs.filter(function(run) { + return run.result === 'pass' && + metric_case_keys.every(key => run[key] === selected[key]) && + (!selected_branch || + nipa_br_pfx_get(run.branch) === selected_branch); + }); +} + +// Links can name a data file, which must not come from another site +function metric_same_origin(source) +{ + try { + return new URL(source, window.location.href).origin === + window.location.origin; + } catch (error) { + return false; + } +} + +function metric_query_url(params) +{ + const query = new URLSearchParams(); + for (const key of metric_case_keys) { + if (params.get(key)) + query.set(key, params.get(key)); + } + if (params.get('branch')) + query.set('br-pfx', params.get('branch')); + return `query/metrics?${query.toString()}`; +} + +function metric_summary_label(metric) +{ + const name = metric.transform === 'rate' ? + `${metric.name} rate` : metric.name; + if (metric.aggregation) + return `${name} (${metric.aggregation}, ` + + `${metric.summarize.toUpperCase()})`; + if (metric.summarize !== 'last') + return `${name} (${metric.summarize.toUpperCase()})`; + return name; +} + +function summary_metric_rows(runs) +{ + const rows = []; + + for (const run of runs) { + for (const metric of run.short_json) { + const aggregated = Boolean(metric.aggregation); + const value = metric.summary[metric.summarize]; + rows.push({ + ...run, + metric: metric.name, + metric_label: metric_summary_label(metric), + host: metric.host || '', + origin: metric.origin || '', + display_range: aggregated ? {} : metric.display_range, + display_label: aggregated ? null : metric.display_label, + summary: metric.summarize, + samples: metric.samples, + value: aggregated ? value : + value * (metric.display_scale ?? 1), + }); + } + } + return rows; +} + +function series_label(rows, row) +{ + const origins = new Set(rows.map(entry => entry.origin)); + return [origins.size > 1 ? row.origin : '', row.host] + .filter(Boolean).join(' / '); +} + +function metric_groups(rows) +{ + const groups = new Map(); + + for (const row of rows) { + const key = row.metric_label; + if (!groups.has(key)) + groups.set(key, []); + groups.get(key).push(row); + } + + return [...groups.values()].sort((a, b) => + a[0].metric.localeCompare(b[0].metric) || + a[0].metric_label.localeCompare(b[0].metric_label)); +} + +function metric_host_compare(a, b) +{ + const order = {local: 0, remote: 1}; + const a_order = order[a] ?? 2; + const b_order = order[b] ?? 2; + return a_order - b_order || a.localeCompare(b); +} + +function metric_detail_url(row) +{ + const current = new URLSearchParams(window.location.search); + const params = new URLSearchParams({id: row.id}); + const overview_data = current.get('data') || metric_data_source; + const detail_data = current.get('detail-data') || + overview_data?.replace(/result-metrics\.json$/, + 'result-metric-details.json'); + if (detail_data) + params.set('data', detail_data); + return `metric-details.html?${params.toString()}`; +} + +function render_one_chart(container, rows) +{ + const labels = [...new Set(rows.map(row => row.branch))]; + const branch_dates = new Map(rows.map(row => + [row.branch, row.branch_date])); + const series = new Map(); + for (const row of rows) { + const key = [row.origin, row.host].join('\x1f'); + if (!series.has(key)) + series.set(key, new Map()); + series.get(key).set(row.branch, row); + } + + const datasets = []; + let color_index = 0; + const sorted_series = [...series.values()].sort(function(a, b) { + const first_a = a.values().next().value; + const first_b = b.values().next().value; + return first_a.origin.localeCompare(first_b.origin) || + metric_host_compare(first_a.host, first_b.host); + }); + for (const points of sorted_series) { + const first = points.values().next().value; + const color = chart_colors[color_index++ % chart_colors.length]; + const point_rows = labels.map(branch => points.get(branch) || null); + datasets.push({ + label: series_label(rows, first), + data: point_rows.map(row => row ? row.value : null), + metricRows: point_rows, + borderColor: color, + backgroundColor: color, + borderWidth: 2, + pointRadius: 3, + pointHoverRadius: 6, + pointHitRadius: 12, + spanGaps: false, + tension: 0, + }); + } + + const metric_label = rows[0].metric_label; + const display_label = rows[0].display_label; + const display_range = rows[0].display_range || {}; + const box = document.createElement('div'); + box.className = 'metric-chart-box'; + const canvas = document.createElement('canvas'); + box.appendChild(canvas); + container.appendChild(box); + + new Chart(canvas, { + type: 'line', + data: {labels: labels, datasets: datasets}, + options: { + responsive: true, + maintainAspectRatio: false, + interaction: {mode: 'nearest', intersect: true}, + plugins: { + legend: {position: 'bottom'}, + title: {display: true, text: metric_label}, + tooltip: { + callbacks: { + label: function(context) { + const row = context.dataset.metricRows[context.dataIndex]; + return `${context.dataset.label}: ${row.value}`; + }, + afterBody: function(items) { + const item = items[0]; + const row = item.dataset.metricRows[item.dataIndex]; + return [ + `Summary: ${row.summary}`, + `Samples: ${row.samples}`, + `Branch: ${row.branch}`, + `Run: ${new Date(row.start).toLocaleString()}`, + ]; + }, + }, + }, + }, + scales: { + x: { + title: {display: true, text: 'Branch'}, + ticks: { + callback: function(value) { + return branch_dates.get( + this.getLabelForValue(value)); + }, + maxRotation: 60, + minRotation: 30, + }, + }, + y: { + min: display_range.min, + max: display_range.max, + title: { + display: true, + text: display_label || metric_label, + }, + }, + }, + onClick: function(event, points, chart) { + if (!points.length) + return; + const point = points[0]; + const row = chart.data.datasets[point.datasetIndex] + .metricRows[point.index]; + window.location.href = metric_detail_url(row); + }, + }, + }); +} + +function render_metric_charts() +{ + const runs = filtered_runs(); + const rows = summary_metric_rows(runs); + const status = document.getElementById('metric-status'); + const container = document.getElementById('metric-charts'); + + container.replaceChildren(); + + if (!runs.length) { + status.hidden = false; + status.textContent = metric_runs.length ? + 'No measurements match these filters.' : + 'No performance measurements have been collected yet.'; + return; + } + + const groups = metric_groups(rows); + for (const group of groups) + render_one_chart(container, group); + + status.textContent = ''; + status.hidden = true; +} + +async function load_metrics() +{ + const status = document.getElementById('metric-status'); + status.hidden = false; + status.textContent = 'Loading...'; + + const url_params = new URLSearchParams(window.location.search); + metric_data_source = url_params.get('data'); + if (metric_data_source && !metric_same_origin(metric_data_source)) { + status.textContent = 'Not loading measurements from another site: ' + + metric_data_source; + return; + } + // Static sources hold every test case, the backend can select one + const url = metric_data_source || metric_query_url(url_params); + + try { + const response = await fetch(url); + if (!response.ok) + throw new Error(`HTTP ${response.status}`); + metric_runs = await response.json(); + const selected = selected_case(url_params); + const title = document.getElementById('metric-title'); + if (selected) { + title.textContent = + `${case_name(selected)} - ${runner_name(selected)}`; + } + render_metric_charts(); + } catch (error) { + status.hidden = false; + status.textContent = `Failed to load measurements from ${url}: ${error}`; + } +} + +function do_it() +{ + load_metrics(); +} From ba59774402a159d6ad1642cdbbd9b6f360753a52 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:21:23 -0700 Subject: [PATCH 10/18] ui: add sampled performance metric details Add a per-run page for complete timestamped observations. Apply producer transforms and display metadata, and show regression baseline information. Load a data file named in the URL only from this site, and link to test outputs only over HTTP(S), as they come from remote runners. Signed-off-by: Stanislav Fomichev --- ui/metric-details.html | 57 ++++++ ui/metric-details.js | 399 +++++++++++++++++++++++++++++++++++++++++ 2 files changed, 456 insertions(+) create mode 100644 ui/metric-details.html create mode 100644 ui/metric-details.js diff --git a/ui/metric-details.html b/ui/metric-details.html new file mode 100644 index 00000000..67663e36 --- /dev/null +++ b/ui/metric-details.html @@ -0,0 +1,57 @@ + + + + + NIPA metric details + + + + + + + + + + + +
+

Metric details

+

Loading...

+ + + + + diff --git a/ui/metric-details.js b/ui/metric-details.js new file mode 100644 index 00000000..0691861e --- /dev/null +++ b/ui/metric-details.js @@ -0,0 +1,399 @@ +let metric_result = null; + +function select_metric_result(payload, id) +{ + return Array.isArray(payload) ? + payload.find(entry => String(entry.id) === id) : payload; +} + +function metric_host_compare(a, b) +{ + const order = {local: 0, remote: 1}; + const a_order = order[a || ''] ?? 2; + const b_order = order[b || ''] ?? 2; + return a_order - b_order || (a || '').localeCompare(b || ''); +} + +function metric_sort(a, b) +{ + if (a.name !== b.name) + return a.name.localeCompare(b.name); + return metric_host_compare(a.host, b.host); +} + +// Links can name a data file, which must not come from another site +function metric_same_origin(source) +{ + try { + return new URL(source, window.location.href).origin === + window.location.origin; + } catch (error) { + return false; + } +} + +// Results come from remote runners, follow their web links only +function metric_web_link(link) +{ + if (!link) + return null; + try { + const url = new URL(link, window.location.href); + return ['http:', 'https:'].includes(url.protocol) ? url.href : null; + } catch (error) { + return null; + } +} + +function add_metadata_row(table, label, value, link = null) +{ + const row = table.insertRow(); + const heading = document.createElement('th'); + heading.scope = 'row'; + heading.textContent = label; + row.appendChild(heading); + const cell = row.insertCell(); + if (link) { + const anchor = document.createElement('a'); + anchor.href = link; + anchor.textContent = value; + cell.appendChild(anchor); + } else { + cell.textContent = value; + } + return cell; +} + +function render_value_metadata(table) +{ + // Metrics of the selftests have one value, rather than samples to chart + for (const metric of metric_result.short_json) { + if (!metric.goal) + continue; + + const host = metric.host ? ` / ${metric.host}` : ''; + const unit = metric.display_label ? ` ${metric.display_label}` : ''; + add_metadata_row(table, `${metric.name}${host}`, + `${metric.summary[metric.summarize]}${unit}`); + } +} + +function render_regression_metadata(table) +{ + for (const metric of metric_result.short_json) { + const result = metric.regression_result; + if (!result) + continue; + + const host = metric.host ? ` / ${metric.host}` : ''; + let text; + if (result.status === 'insufficient-data') { + text = `Insufficient history (${result.baseline_samples} samples)`; + } else if (result.status === 'error') { + text = `error: ${result.current}, must be zero`; + } else if (result.baseline === undefined) { + text = `${result.status}: ${result.current}`; + } else { + const delta = result.delta_percent === undefined ? '' : + `, ${result.delta_percent.toFixed(1)}%`; + text = `${result.status}: ${result.current} versus ` + + `${result.baseline}${delta}`; + } + const cell = add_metadata_row( + table, `Regression (${metric.name}${host})`, text); + if (result.status === 'regression' || result.status === 'error') { + cell.style.backgroundColor = '#d06060'; + cell.style.color = 'white'; + } + } +} + +function render_metadata() +{ + const table = document.getElementById('metric-metadata'); + table.replaceChildren(); + add_metadata_row(table, 'Branch', metric_result.branch); + add_metadata_row(table, 'Runner', + `${metric_result.remote}:${metric_result.executor}`); + add_metadata_row(table, 'Test case', + `${metric_result.group}:${metric_result.test}:` + + metric_result.subtest); + add_metadata_row(table, 'Started', + new Date(metric_result.start).toLocaleString()); + add_metadata_row(table, 'Result', metric_result.result); + add_metadata_row(table, 'Attempt', metric_result.attempt.toString()); + render_value_metadata(table); + render_regression_metadata(table); + const output = metric_web_link(metric_result.link); + if (output) + add_metadata_row(table, 'Output', 'Open test output', output); + table.hidden = false; + document.title = `${metric_result.test}:${metric_result.subtest} metrics`; +} + +function metric_dimensions(metric) +{ + const dimensions = {}; + for (const [key, value] of Object.entries(metric)) { + if (!['time', 'value'].includes(key)) + dimensions[key] = value; + } + return dimensions; +} + +function rate_observations(observations) +{ + const rates = []; + for (let index = 1; index < observations.length; index++) { + const previous = observations[index - 1]; + const current = observations[index]; + const elapsed = current.time - previous.time; + if (elapsed <= 0) + continue; + + let values; + if (Array.isArray(current.value)) { + values = current.value.map(function(value, cpu) { + const old_value = previous.value[cpu]; + if (value === null || old_value === null || value < old_value) + return null; + return (value - old_value) / elapsed; + }); + } else { + if (current.value < previous.value) + continue; + values = (current.value - previous.value) / elapsed; + } + rates.push({time: current.time, value: values}); + } + return rates; +} + +function scale_observations(observations, scale) +{ + return observations.map(function(observation) { + const value = Array.isArray(observation.value) ? + observation.value.map(item => item === null ? + null : item * scale) : observation.value * scale; + return {time: observation.time, value: value}; + }); +} + +function sampled_metric_groups() +{ + const groups = new Map(); + for (const metric of metric_result.full_json) { + if (metric.time === undefined) + continue; + const dimensions = metric_dimensions(metric); + const key = JSON.stringify(Object.entries(dimensions).sort()); + if (!groups.has(key)) { + groups.set(key, { + ...dimensions, + observations: [], + }); + } + groups.get(key).observations.push({ + time: metric.time, + value: metric.value, + }); + } + + return [...groups.values()] + .map(function(group) { + group.observations.sort((a, b) => a.time - b.time); + if (group.transform === 'rate') + group.observations = rate_observations(group.observations); + const display_scale = group.display_scale ?? 1; + if (display_scale !== 1) { + group.observations = scale_observations( + group.observations, display_scale); + } + return group; + }) + .filter(group => group.observations.length) + .sort(metric_sort); +} + +function series_color(index, count) +{ + const hue = Math.round(index * 360 / count); + return `hsl(${hue}, 70%, 48%)`; +} + +function per_cpu_datasets(group) +{ + let cpu_count = 0; + for (const observation of group.observations) + cpu_count = Math.max(cpu_count, observation.value.length); + + const values_by_cpu = []; + for (let cpu = 0; cpu < cpu_count; cpu++) { + const values = group.observations.map( + observation => observation.value[cpu] ?? null); + const present = values.filter(value => value !== null); + if (present.length) + values_by_cpu.push({cpu: cpu, values: values, + peak: Math.max(...present)}); + } + values_by_cpu.sort((a, b) => b.peak - a.peak || a.cpu - b.cpu); + + return { + total: values_by_cpu.length, + datasets: values_by_cpu.map(function(series) { + // Keep a CPU's color stable across every metric and host chart. + const color = series_color(series.cpu, cpu_count); + return { + label: `CPU ${series.cpu}`, + data: series.values, + borderColor: color, + backgroundColor: color, + borderWidth: 1.5, + pointRadius: 1, + pointHoverRadius: 5, + spanGaps: false, + tension: 0, + }; + }), + }; +} + +function scalar_datasets(group) +{ + return { + total: 1, + datasets: [{ + label: group.host || group.name, + data: group.observations.map(observation => observation.value), + borderColor: '#3366cc', + backgroundColor: '#3366cc', + borderWidth: 2, + pointRadius: 3, + pointHoverRadius: 6, + spanGaps: false, + tension: 0, + }], + }; +} + +function render_sampled_chart(container, group) +{ + const series = group.type === 'per-cpu' ? + per_cpu_datasets(group) : scalar_datasets(group); + if (!series.datasets.length) + return false; + + const box = document.createElement('div'); + box.className = 'metric-chart-box'; + const canvas = document.createElement('canvas'); + box.appendChild(canvas); + container.appendChild(box); + + const dimensions = [group.origin, group.host].filter(Boolean).join(' / '); + const dimension_label = dimensions ? ` / ${dimensions}` : ''; + const cpu_count = group.type === 'per-cpu' ? + ` (${series.total} CPUs)` : ''; + const metric_name = group.transform === 'rate' ? + `${group.name} rate` : group.name; + const display_range = group.display_range || {}; + const title = `${metric_name}${dimension_label}${cpu_count}`; + new Chart(canvas, { + type: 'line', + data: { + labels: group.observations.map(observation => observation.time), + datasets: series.datasets, + }, + options: { + responsive: true, + maintainAspectRatio: false, + interaction: {mode: 'nearest', intersect: false}, + plugins: { + legend: { + display: series.total <= 32, + position: 'bottom', + }, + title: { + display: true, + text: title, + }, + tooltip: { + callbacks: { + afterBody: function(items) { + const time = items[0].label; + return `Time: ${time} seconds into test`; + }, + }, + }, + }, + scales: { + x: {title: {display: true, text: 'Seconds into test'}}, + y: { + min: display_range.min, + max: display_range.max, + title: { + display: true, + text: group.display_label || metric_name, + }, + }, + }, + }, + }); + return true; +} + +function render_sampled_charts() +{ + const container = document.getElementById('sampled-charts'); + container.replaceChildren(); + const groups = sampled_metric_groups(); + let rendered = 0; + for (const group of groups) { + if (render_sampled_chart(container, group)) + rendered++; + } + document.getElementById('sample-section').hidden = rendered === 0; + return rendered; +} + +async function load_metric_details() +{ + const status = document.getElementById('metric-status'); + const params = new URLSearchParams(window.location.search); + const data_source = params.get('data'); + const id = params.get('id'); + let url; + if (data_source && !metric_same_origin(data_source)) { + status.hidden = false; + status.textContent = 'Not loading metric details from another ' + + `site: ${data_source}`; + return; + } else if (data_source) { + url = data_source; + } else if (id) { + url = `query/metrics/${encodeURIComponent(id)}`; + } else { + status.hidden = false; + status.textContent = 'No metric result was selected.'; + return; + } + + try { + const response = await fetch(url); + if (!response.ok) + throw new Error(`HTTP ${response.status}`); + const payload = await response.json(); + metric_result = select_metric_result(payload, id); + render_metadata(); + render_sampled_charts(); + status.textContent = ''; + status.hidden = true; + } catch (error) { + status.hidden = false; + status.textContent = `Failed to load metric details from ${url}: ${error}`; + } +} + +function do_it() +{ + load_metric_details(); +} From 140dc426528d493d867934231caeaa18263b4058 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 16 Sep 2026 08:21:28 -0700 Subject: [PATCH 11/18] ui: add performance results table Add a date-oriented performance table with branch, runner, test, and regression filters. Link test cases to history and results to observations. Load a data file named in the URL only from this site. Signed-off-by: Stanislav Fomichev --- ui/metric-table.html | 174 ++++++++++++++++++ ui/metric-table.js | 414 +++++++++++++++++++++++++++++++++++++++++++ ui/mock/README.rst | 24 +++ ui/sitemap.html | 3 +- 4 files changed, 614 insertions(+), 1 deletion(-) create mode 100644 ui/metric-table.html create mode 100644 ui/metric-table.js create mode 100644 ui/mock/README.rst diff --git a/ui/metric-table.html b/ui/metric-table.html new file mode 100644 index 00000000..99d542f0 --- /dev/null +++ b/ui/metric-table.html @@ -0,0 +1,174 @@ + + + + + NIPA performance metrics + + + + + + + + + + +
+

Performance metrics

+
+ Filtering +
+
+ + + + +
+
+
+
+
+ +
+
+
+ Sort +
+ +
+

+ Update URL

+
+
+
+

Loading...

+

+ within baseline + regression + insufficient history + not tracked +

+
+ +
+ + diff --git a/ui/metric-table.js b/ui/metric-table.js new file mode 100644 index 00000000..7473e14a --- /dev/null +++ b/ui/metric-table.js @@ -0,0 +1,414 @@ +let metric_table_runs = []; +let metric_table_data_source = null; + +function metric_runner_name(run) +{ + return `${run.remote}:${run.executor}`; +} + +function metric_case_name(run) +{ + return `${run.group}:${run.test}:${run.subtest}`; +} + +function metric_run_date(run) +{ + return run.branch_date.slice(0, 10); +} + +function metric_detail_data_source() +{ + const params = new URLSearchParams(window.location.search); + const source = params.get('detail-data'); + if (source) + return source; + if (metric_table_data_source?.endsWith('result-metrics.json')) + return metric_table_data_source.replace(/result-metrics\.json$/, + 'result-metric-details.json'); + return null; +} + +function metric_detail_url(run) +{ + const params = new URLSearchParams(); + const source = metric_detail_data_source(); + + params.set('id', run.id); + if (source) + params.set('data', source); + return `metric-details.html?${params.toString()}`; +} + +function metric_history_url(run) +{ + const params = new URLSearchParams(); + for (const key of ['remote', 'executor', 'group', 'test', 'subtest']) + params.set(key, run[key]); + params.set('branch', nipa_br_pfx_get(run.branch)); + if (metric_table_data_source) + params.set('data', metric_table_data_source); + const source = metric_detail_data_source(); + if (source) + params.set('detail-data', source); + return `metrics.html?${params.toString()}`; +} + +function metric_run_is_newer(candidate, current) +{ + if (!current) + return true; + if (candidate.result === 'pass' && current.result !== 'pass') + return true; + if (candidate.result !== 'pass' && current.result === 'pass') + return false; + + if (candidate.start !== current.start) + return candidate.start > current.start; + return candidate.id > current.id; +} + +function metric_regression_results(run) +{ + return run.short_json + .filter(metric => metric.regression_result) + .map(metric => ({ + name: metric.name, + host: metric.host || '', + display_label: metric.aggregation || metric.display_label || '', + ...metric.regression_result, + })); +} + +function metric_result_status(run) +{ + if (run.result !== 'pass') + return 'regression'; + + const regressions = metric_regression_results(run); + if (regressions.some(result => result.status === 'regression' || + result.status === 'error')) + return 'regression'; + if (regressions.some(result => result.status === 'insufficient-data')) + return 'insufficient'; + if (regressions.some(result => result.status === 'ok')) + return 'ok'; + return 'untracked'; +} + +function metric_result_class(run) +{ + const suffix = { + regression: '', + insufficient: '-insufficient', + ok: '-ok', + untracked: '-untracked', + }[metric_result_status(run)]; + return `metric-regression${suffix}`; +} + +function metric_regression_tooltip(run) +{ + const lines = []; + for (const result of metric_regression_results(run)) { + const label = result.host ? `${result.name} / ${result.host}` : + result.name; + const unit = result.display_label ? ` ${result.display_label}` : ''; + if (result.status === 'insufficient-data') { + lines.push(`${label}: insufficient history ` + + `(${result.baseline_samples} samples, ` + + `current ${result.current}${unit})`); + } else if (result.status === 'error' || + result.baseline === undefined) { + lines.push(`${label}: ${result.status} ` + + `(current ${result.current}${unit}, must be zero)`); + } else { + const change = result.delta_percent === undefined ? '' : + `, ${result.delta_percent.toFixed(1)}%`; + lines.push(`${label}: ${result.status}${change} ` + + `(current ${result.current}${unit}, ` + + `baseline ${result.baseline}${unit})`); + } + } + return lines; +} + +function populate_metric_select(id, values) +{ + const select = document.getElementById(id); + select.replaceChildren(new Option('-- all --', '')); + for (const value of [...values].sort()) + select.add(new Option(value, value)); +} + +function populate_metric_filters() +{ + populate_metric_select('branch-filter', new Set( + metric_table_runs.map(run => nipa_br_pfx_get(run.branch)))); + populate_metric_select('runner-filter', new Set( + metric_table_runs.map(metric_runner_name))); + document.getElementById('date-count').value = '30'; + document.getElementById('name-filter').value = ''; + for (const id of ['show-regression', 'show-ok', 'show-insufficient', + 'show-untracked']) + document.getElementById(id).checked = true; + document.getElementById('sort-name').checked = true; + + const params = new URLSearchParams(window.location.search); + const values = { + 'date-count': params.get('date-count'), + 'branch-filter': params.get('branch-filter'), + 'runner-filter': params.get('runner-filter'), + 'name-filter': params.get('name-filter'), + }; + for (const [id, value] of Object.entries(values)) { + if (value !== null) + document.getElementById(id).value = value; + } + for (const id of ['show-regression', 'show-ok', 'show-insufficient', + 'show-untracked']) { + if (params.get(id) === '0') + document.getElementById(id).checked = false; + } + if (params.get('sort') === 'regression') + document.getElementById('sort-regression').checked = true; + + document.getElementById('metric-controls').disabled = false; +} + +function metric_status_is_visible(status) +{ + return document.getElementById(`show-${status}`).checked; +} + +function update_metric_table_url() +{ + const url = new URL(window.location.href); + for (const id of ['date-count', 'branch-filter', 'runner-filter', + 'name-filter', 'show-regression', 'show-ok', + 'show-insufficient', 'show-untracked']) + url.searchParams.delete(id); + url.searchParams.delete('sort'); + + const date_count = document.getElementById('date-count').value; + const branch = document.getElementById('branch-filter').value; + const runner = document.getElementById('runner-filter').value; + const name = document.getElementById('name-filter').value; + if (date_count !== '30') + url.searchParams.set('date-count', date_count); + if (branch) + url.searchParams.set('branch-filter', branch); + if (runner) + url.searchParams.set('runner-filter', runner); + if (name) + url.searchParams.set('name-filter', name); + for (const id of ['show-regression', 'show-ok', 'show-insufficient', + 'show-untracked']) { + if (!document.getElementById(id).checked) + url.searchParams.set(id, '0'); + } + if (document.getElementById('sort-regression').checked) + url.searchParams.set('sort', 'regression'); + window.history.pushState({}, '', url); +} + +function build_metric_table() +{ + const status = document.getElementById('metric-status'); + const table = document.getElementById('metric-table'); + const dates = new Set(); + const rows = new Map(); + const selected_branch = document.getElementById('branch-filter').value; + const selected_runner = document.getElementById('runner-filter').value; + const needle = document.getElementById('name-filter').value; + let needle_re = null; + let warning = ''; + if (needle) { + try { + needle_re = new RegExp(needle); + } catch (error) { + warning = `Bad test case regex: ${error.message}`; + } + } + + for (const run of metric_table_runs) { + const date = metric_run_date(run); + const branch = nipa_br_pfx_get(run.branch); + const runner = metric_runner_name(run); + const testcase = metric_case_name(run); + if ((selected_branch && branch !== selected_branch) || + (selected_runner && runner !== selected_runner) || + (needle_re && !needle_re.test(testcase))) + continue; + + const key = [branch, runner, testcase].join('\x1f'); + if (!rows.has(key)) + rows.set(key, {branch, runner, testcase, runs: new Map()}); + + const row = rows.get(key); + const current = row.runs.get(date); + if (metric_run_is_newer(run, current)) + row.runs.set(date, run); + } + + for (const [key, row] of rows) { + for (const [date, run] of row.runs) { + if (!metric_status_is_visible(metric_result_status(run))) + row.runs.delete(date); + else + dates.add(date); + } + if (!row.runs.size) + rows.delete(key); + } + + table.replaceChildren(); + if (!rows.size || !dates.size) { + table.hidden = true; + status.hidden = false; + status.textContent = warning || (metric_table_runs.length ? + 'No measurements match these filters.' : + 'No performance measurements have been collected yet.'); + return; + } + + const date_count = Number.parseInt( + document.getElementById('date-count').value, 10) || 30; + const sorted_dates = [...dates].sort().reverse().slice(0, date_count); + const status_order = { + regression: 0, insufficient: 1, untracked: 2, ok: 3, + }; + const sorted_rows = [...rows.values()] + .filter(row => sorted_dates.some(date => row.runs.has(date))) + .sort(function(a, b) { + if (document.getElementById('sort-regression').checked) { + const a_status = Math.min(...sorted_dates + .map(date => a.runs.get(date)) + .filter(Boolean) + .map(run => status_order[metric_result_status(run)])); + const b_status = Math.min(...sorted_dates + .map(date => b.runs.get(date)) + .filter(Boolean) + .map(run => status_order[metric_result_status(run)])); + if (a_status !== b_status) + return a_status - b_status; + } + return a.testcase.localeCompare(b.testcase) || + a.runner.localeCompare(b.runner) || + a.branch.localeCompare(b.branch); + }); + + const head = table.createTHead().insertRow(); + const case_header = document.createElement('th'); + case_header.textContent = 'Test case'; + head.appendChild(case_header); + for (const date of sorted_dates) { + const header = document.createElement('th'); + header.className = 'metric-date'; + header.textContent = date; + head.appendChild(header); + } + + const body = table.createTBody(); + for (const row_data of sorted_rows) { + const row = body.insertRow(); + const name = row.insertCell(); + name.className = 'metric-case'; + + const representative = row_data.runs.values().next().value; + const link = document.createElement('a'); + link.href = metric_history_url(representative); + link.textContent = row_data.testcase; + name.appendChild(link); + + const dimensions = document.createElement('span'); + dimensions.className = 'metric-dimensions'; + dimensions.textContent = `${row_data.runner} · ${row_data.branch}`; + name.appendChild(dimensions); + + for (const date of sorted_dates) { + const cell = row.insertCell(); + const run = row_data.runs.get(date); + if (!run) + continue; + + cell.className = `metric-result ${metric_result_class(run)}`; + const result_link = document.createElement('a'); + result_link.href = metric_detail_url(run); + result_link.textContent = '●'; + const series_count = run.short_json.length; + result_link.title = [ + run.branch, + run.start, + `${run.result}; ${series_count} metric series`, + ...metric_regression_tooltip(run), + ].filter(Boolean).join('\n'); + result_link.setAttribute('aria-label', + `${row_data.testcase} on ${date}: ${run.result}`); + cell.appendChild(result_link); + } + } + + status.hidden = !warning; + status.textContent = warning; + table.hidden = false; +} + +// Links can name a data file, which must not come from another site +function metric_same_origin(source) +{ + try { + return new URL(source, window.location.href).origin === + window.location.origin; + } catch (error) { + return false; + } +} + +async function load_metric_table() +{ + const status = document.getElementById('metric-status'); + const params = new URLSearchParams(window.location.search); + metric_table_data_source = params.get('data'); + if (metric_table_data_source && + !metric_same_origin(metric_table_data_source)) { + status.hidden = false; + status.textContent = 'Not loading measurements from another site: ' + + metric_table_data_source; + return; + } + const url = metric_table_data_source || 'query/metrics'; + + try { + const response = await fetch(url); + if (!response.ok) + throw new Error(`HTTP ${response.status}`); + metric_table_runs = await response.json(); + populate_metric_filters(); + build_metric_table(); + } catch (error) { + status.hidden = false; + status.textContent = `Failed to load measurements from ${url}: ${error}`; + } +} + +function do_it() +{ + for (const id of ['date-count', 'branch-filter', 'runner-filter', + 'show-regression', 'show-ok', 'show-insufficient', + 'show-untracked', 'sort-name', 'sort-regression']) { + document.getElementById(id).addEventListener( + 'change', build_metric_table); + } + document.getElementById('name-filter').addEventListener( + 'input', build_metric_table); + document.getElementById('update-url-button').addEventListener( + 'click', function(event) { + event.preventDefault(); + update_metric_table_url(); + }); + window.addEventListener('popstate', function() { + populate_metric_filters(); + build_metric_table(); + }); + load_metric_table(); +} diff --git a/ui/mock/README.rst b/ui/mock/README.rst new file mode 100644 index 00000000..0da7f847 --- /dev/null +++ b/ui/mock/README.rst @@ -0,0 +1,24 @@ +Result metrics UI fixtures +========================== + +Serve the ``ui`` directory:: + + cd ui + python3 -m http.server 8000 + +Then open:: + + http://127.0.0.1:8000/metric-table.html?data=mock/result-metrics.json + +The ``data`` parameter replaces the ``query/metrics`` backend, and the table +derives the matching ``result-metric-details.json`` path from it. Test-case +links open metric history; result links open complete observations. Filters +can be saved in the URL. + +Without ``data``, the pages load ``query/metrics``, which +``scripts/ui_assets.sh download`` fetches from production for offline use. +Result details still need the backend. + +``result-metrics.json`` models the compact ``/query/metrics`` response. +``result-metric-details.json`` adds the original timestamped observations. +The fixtures include devmem and iperf histories with regression examples. diff --git a/ui/sitemap.html b/ui/sitemap.html index 2ce877ab..76aca416 100644 --- a/ui/sitemap.html +++ b/ui/sitemap.html @@ -3,5 +3,6 @@ Result log | Check stats | Flaky tests | - Device results + Device results | + Performance From a8cc6e1bc35dc7a6bb1ba87ca87560cb289ddaa5 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:47:14 -0700 Subject: [PATCH 12/18] contest: hw: share result parsing Move hardware result parsing out of the deployer so service and local tools can use it without loading service dependencies. Signed-off-by: Stanislav Fomichev --- contest/hw/lib/deployer.py | 137 +------------------------------ contest/hw/lib/result_parser.py | 141 ++++++++++++++++++++++++++++++++ 2 files changed, 144 insertions(+), 134 deletions(-) create mode 100644 contest/hw/lib/result_parser.py diff --git a/contest/hw/lib/deployer.py b/contest/hw/lib/deployer.py index 9d8accb4..fe942559 100644 --- a/contest/hw/lib/deployer.py +++ b/contest/hw/lib/deployer.py @@ -13,8 +13,9 @@ import time from dataclasses import dataclass, field -from lib.nipa import (has_crash, extract_crash, guess_indicators, - result_from_indicators, parse_nested_tests, namify) +from lib.nipa import has_crash, extract_crash +# hwksft and its tests import parse_results() from here +from lib.result_parser import parse_results # noqa: F401 # Log file handle, set by set_log_file() before builds start. @@ -528,138 +529,6 @@ def read_device_info(results_path): return json.load(fp) -def parse_results(results_path, link): - """Parse fetched test output into a vmksft-p-style result list. - - Reads info/stdout files from the test-outputs directory and - the .attempted file to identify crashed tests. - """ - output_dir = os.path.join(results_path, 'test-outputs') - - # Output dirs with no 'info' belong either to a retry or to a test which - # never got to write one because the machine died under it. Remember the - # latter by test name, the .attempted pass below has no other way of - # finding whatever output did make it to disk. Names are only unique - # within a target, so drop the ones we'd have to guess about. - orphan_links = {} - - # Parse each test output directory - cases = [] - completed_tests = set() - if os.path.isdir(output_dir): - for entry in sorted(os.listdir(output_dir)): - test_dir = os.path.join(output_dir, entry) - if not os.path.isdir(test_dir): - continue - - info_path = os.path.join(test_dir, 'info') - stdout_path = os.path.join(test_dir, 'stdout') - entry_link = f'{link}/test-outputs/{entry}/' - - if not os.path.exists(info_path): - if not entry.endswith('-retry'): - # hw_worker names the dirs "$index-$namify(prog)" - name = entry.split('-', 1)[-1] - if name in orphan_links: - orphan_links[name] = None - else: - orphan_links[name] = entry_link - continue - - try: - with open(info_path, encoding='utf-8') as fp: - info = json.load(fp) - except (json.JSONDecodeError, ValueError): - print(f"parse_results: bad JSON in {info_path}, marking as fail") - cases.append({ - 'test': entry, - 'group': 'selftests-hw', - 'result': 'fail', - 'link': entry_link, - }) - continue - - retcode = info.get('retcode', 1) - target = info.get('target', 'unknown') - prog = info.get('prog', entry) - test_name = f"{target}:{prog}" - completed_tests.add(test_name) - - stdout = '' - if os.path.exists(stdout_path): - with open(stdout_path, encoding='utf-8') as fp: - stdout = fp.read() - - # Determine result using indicators - indicators = guess_indicators(stdout) - result = result_from_indicators(retcode, indicators) - - # Parse nested subtests - nested = parse_nested_tests(stdout, namify) - - # Determine retry result if present - retry_result = None - retry_nested = None - if 'retry_retcode' in info: - retry_stdout = '' - retry_dir = os.path.join(output_dir, entry + '-retry') - retry_stdout_path = os.path.join(retry_dir, 'stdout') - if os.path.exists(retry_stdout_path): - with open(retry_stdout_path, encoding='utf-8') as fp: - retry_stdout = fp.read() - retry_indicators = guess_indicators(retry_stdout) - retry_result = result_from_indicators(info['retry_retcode'], - retry_indicators) - if nested: - retry_nested = list(nested) - parse_nested_tests(retry_stdout, namify, - prev_results=retry_nested) - - safe_name = namify(prog) - - outcome = { - 'test': safe_name or entry, - 'group': f'selftests-{namify(target)}', - 'result': result, - 'link': entry_link, - } - if 'time' in info: - outcome['time'] = info['time'] - if retry_result is not None: - outcome['retry'] = retry_result - if info.get('crashes'): - outcome['crashes'] = info['crashes'] - if info.get('warnings'): - outcome['warnings'] = info['warnings'] - if retry_nested: - outcome['results'] = retry_nested - elif nested: - outcome['results'] = nested - cases.append(outcome) - - # Check .attempted for crashed tests (attempted but no output) - attempted_path = os.path.join(results_path, 'attempted.json') - if os.path.exists(attempted_path): - with open(attempted_path, encoding='utf-8') as fp: - try: - attempted = json.load(fp) - except (json.JSONDecodeError, ValueError): - attempted = [] - - for test_name in attempted: - if test_name not in completed_tests: - prog = test_name.split(':', 1)[-1] - cases.append({ - 'test': test_name, - 'group': 'selftests-hw', - 'result': 'fail', - 'link': orphan_links.get(namify(prog)) or link, - 'crashes': ['kernel crash during test'], - }) - - return cases - - def process_crashes(results_path, tree_path, filters): """Post-process crash data from test output directories. diff --git a/contest/hw/lib/result_parser.py b/contest/hw/lib/result_parser.py new file mode 100644 index 00000000..e9484a82 --- /dev/null +++ b/contest/hw/lib/result_parser.py @@ -0,0 +1,141 @@ +# SPDX-License-Identifier: GPL-2.0 + +"""Parse HW worker artifacts without requiring the NIPA service stack.""" + +import json +import os + +from .nipa import guess_indicators, namify, parse_nested_tests, \ + result_from_indicators + + +def parse_results(results_path, link): + """Parse fetched test output into a vmksft-p-style result list. + + Reads info/stdout files from the test-outputs directory and + the .attempted file to identify crashed tests. + """ + output_dir = os.path.join(results_path, 'test-outputs') + + # Output dirs with no 'info' belong either to a retry or to a test which + # never got to write one because the machine died under it. Remember the + # latter by test name, the .attempted pass below has no other way of + # finding whatever output did make it to disk. Names are only unique + # within a target, so drop the ones we'd have to guess about. + orphan_links = {} + + # Parse each test output directory + cases = [] + completed_tests = set() + if os.path.isdir(output_dir): + for entry in sorted(os.listdir(output_dir)): + test_dir = os.path.join(output_dir, entry) + if not os.path.isdir(test_dir): + continue + + info_path = os.path.join(test_dir, 'info') + stdout_path = os.path.join(test_dir, 'stdout') + entry_link = f'{link}/test-outputs/{entry}/' + + if not os.path.exists(info_path): + if not entry.endswith('-retry'): + # hw_worker names the dirs "$index-$namify(prog)" + name = entry.split('-', 1)[-1] + if name in orphan_links: + orphan_links[name] = None + else: + orphan_links[name] = entry_link + continue + + try: + with open(info_path, encoding='utf-8') as fp: + info = json.load(fp) + except (json.JSONDecodeError, ValueError): + print(f"parse_results: bad JSON in {info_path}, marking as fail") + cases.append({ + 'test': entry, + 'group': 'selftests-hw', + 'result': 'fail', + 'link': entry_link, + }) + continue + + retcode = info.get('retcode', 1) + target = info.get('target', 'unknown') + prog = info.get('prog', entry) + test_name = f"{target}:{prog}" + completed_tests.add(test_name) + + stdout = '' + if os.path.exists(stdout_path): + with open(stdout_path, encoding='utf-8') as fp: + stdout = fp.read() + + # Determine result using indicators + indicators = guess_indicators(stdout) + result = result_from_indicators(retcode, indicators) + + # Parse nested subtests + nested = parse_nested_tests(stdout, namify) + + # Determine retry result if present + retry_result = None + retry_nested = None + if 'retry_retcode' in info: + retry_stdout = '' + retry_dir = os.path.join(output_dir, entry + '-retry') + retry_stdout_path = os.path.join(retry_dir, 'stdout') + if os.path.exists(retry_stdout_path): + with open(retry_stdout_path, encoding='utf-8') as fp: + retry_stdout = fp.read() + retry_indicators = guess_indicators(retry_stdout) + retry_result = result_from_indicators(info['retry_retcode'], + retry_indicators) + if nested: + retry_nested = list(nested) + parse_nested_tests(retry_stdout, namify, + prev_results=retry_nested) + + safe_name = namify(prog) + + outcome = { + 'test': safe_name or entry, + 'group': f'selftests-{namify(target)}', + 'result': result, + 'link': entry_link, + } + if 'time' in info: + outcome['time'] = info['time'] + if retry_result is not None: + outcome['retry'] = retry_result + if info.get('crashes'): + outcome['crashes'] = info['crashes'] + if info.get('warnings'): + outcome['warnings'] = info['warnings'] + if retry_nested: + outcome['results'] = retry_nested + elif nested: + outcome['results'] = nested + cases.append(outcome) + + # Check .attempted for crashed tests (attempted but no output) + attempted_path = os.path.join(results_path, 'attempted.json') + if os.path.exists(attempted_path): + with open(attempted_path, encoding='utf-8') as fp: + try: + attempted = json.load(fp) + except (json.JSONDecodeError, ValueError): + attempted = [] + + for test_name in attempted: + if test_name not in completed_tests: + prog = test_name.split(':', 1)[-1] + cases.append({ + 'test': test_name, + 'group': 'selftests-hw', + 'result': 'fail', + 'link': orphan_links.get(namify(prog)) or link, + 'crashes': ['kernel crash during test'], + }) + + return cases From bf83f539e91ce3f91d66feeac5178b6278f8ee2f Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:47:35 -0700 Subject: [PATCH 13/18] contest: hw: add system metric collection Sample CPU and NIC counters on the DUT and peer. Write metrics for each measured test-case window the samples cover, and none if the sampler failed, rather than fill the gaps with an idle host. Signed-off-by: Stanislav Fomichev --- contest/hw/lib/system_monitor.py | 375 ++++++++++++++++++++++++ contest/hw/tests/test_system_monitor.py | 105 +++++++ 2 files changed, 480 insertions(+) create mode 100644 contest/hw/lib/system_monitor.py create mode 100644 contest/hw/tests/test_system_monitor.py diff --git a/contest/hw/lib/system_monitor.py b/contest/hw/lib/system_monitor.py new file mode 100644 index 00000000..85dcf9fb --- /dev/null +++ b/contest/hw/lib/system_monitor.py @@ -0,0 +1,375 @@ +# SPDX-License-Identifier: GPL-2.0 + +"""Collect host counters for isolated performance test executions.""" + +import json +import os +import shlex +import subprocess +import threading +import time + + +CPU_FIELDS = ("user", "nice", "sys", "idle", "iowait", "irq", "sirq", + "steal") +CPU_TIME_METRICS = { + "cpu.time.usr": ("user", "nice"), + "cpu.time.sys": ("sys",), + "cpu.time.sirq": ("sirq",), +} +NIC_METRICS = { + "rx_dropped": "nic.rx.dropped", + "rx_missed_errors": "nic.rx.missed_errors", + "tx_dropped": "nic.tx.dropped", +} +CLOCK_TICKS = 100 +SNAPSHOT_END = "__NIPA_SYSTEM_STAT_END__" +NIC_STATS = "__NIPA_NIC_STATS__" + + +class SystemMonitor: + """Sample raw CPU and NIC counters on a DUT and an optional SSH peer.""" + + def __init__(self, ifname, remote=None, remote_ifname=None, + sample_interval=1.0, clock=time.monotonic): + self.ifname = ifname + self.remote = remote + self.remote_ifname = remote_ifname + self.sample_interval = sample_interval + self._clock = clock + self._origin = None + self._remote_sampler = None + self._thread = None + self._stop_event = threading.Event() + self._ready_event = threading.Event() + self._lock = threading.Lock() + self._samples = [] + self.error = None + + @staticmethod + def _parse_cpu(text): + stats = {} + for line in text.splitlines(): + fields = line.split() + if not fields or not fields[0].startswith("cpu") or \ + not fields[0][3:].isdigit(): + continue + stats[int(fields[0][3:])] = { + name: float(value) + for name, value in zip(CPU_FIELDS, fields[1:]) + } + return stats + + @staticmethod + def _parse_nic(text): + for line in text.splitlines(): + if not line.startswith(NIC_STATS): + continue + fields = line.split() + return { + name: float(value) + for name, value in zip(NIC_METRICS, fields[1:]) + } + raise ValueError("remote NIC statistics are missing") + + @staticmethod + def _read_nic(ifname): + stats = {} + for name in NIC_METRICS: + path = f"/sys/class/net/{ifname}/statistics/{name}" + with open(path, encoding="utf-8") as stat_file: + stats[name] = float(stat_file.read()) + return stats + + def elapsed(self): + return self._clock() - self._origin + + def _start_remote_sampler(self): + if not self.remote or not self.remote_ifname: + return + + nic_reads = "" + for name in NIC_METRICS: + path = shlex.quote( + f"/sys/class/net/{self.remote_ifname}/statistics/{name}") + nic_reads += (f"read value < {path} || exit; " + f"printf ' %s' \"$value\"; ") + command = ( + "while :; do cat /proc/stat || exit; " + f"printf '{NIC_STATS}'; {nic_reads}" + f"printf '\\n{SNAPSHOT_END}\\n'; " + f"sleep {shlex.quote(str(self.sample_interval))} || exit; done" + ) + self._remote_sampler = subprocess.Popen( + ["ssh", "-o", "ConnectTimeout=10", + "-o", "StrictHostKeyChecking=no", "-o", "BatchMode=yes", + self.remote, command], + stdout=subprocess.PIPE, stderr=subprocess.DEVNULL, + ) + + def _read_remote(self): + lines = [] + while True: + line = self._remote_sampler.stdout.readline() + if not line: + status = self._remote_sampler.poll() + raise RuntimeError( + f"remote system sampler stopped unexpectedly ({status})") + line = line.decode("utf-8", "replace").rstrip("\n") + if line == SNAPSHOT_END: + return "\n".join(lines) + lines.append(line) + + def _stop_remote_sampler(self): + sampler = self._remote_sampler + self._remote_sampler = None + if sampler is None: + return + sampler.terminate() + try: + sampler.wait(timeout=5) + except subprocess.TimeoutExpired: + sampler.kill() + sampler.wait(timeout=5) + + def _snapshot(self, remote_text=None): + with open("/proc/stat", encoding="utf-8") as stat_file: + local_cpu = self._parse_cpu(stat_file.read()) + hosts = { + "local": { + "cpu": local_cpu, + "nic": self._read_nic(self.ifname), + }, + } + if remote_text is not None: + hosts["remote"] = { + "cpu": self._parse_cpu(remote_text), + "nic": self._parse_nic(remote_text), + } + return {"time": self.elapsed(), "hosts": hosts} + + def _append_sample(self, sample): + with self._lock: + self._samples.append(sample) + self._ready_event.set() + + def _sample_loop(self): + try: + if self._remote_sampler is not None: + while True: + remote_text = self._read_remote() + self._append_sample(self._snapshot(remote_text)) + if self._stop_event.is_set(): + break + else: + while not self._stop_event.is_set(): + self._append_sample(self._snapshot()) + self._stop_event.wait(self.sample_interval) + except Exception as error: # pylint: disable=broad-except + if not self._stop_event.is_set(): + self.error = str(error) + self._ready_event.set() + + def start(self): + """Start sampling and wait for the initial snapshot.""" + self._origin = self._clock() + self._stop_event.clear() + self._ready_event.clear() + self._samples = [] + self.error = None + self._start_remote_sampler() + self._thread = threading.Thread(target=self._sample_loop, daemon=True) + self._thread.start() + if not self._ready_event.wait(timeout=15): + self.stop() + raise TimeoutError("system monitor did not produce an initial sample") + if self.error: + error = self.error + self.stop() + raise RuntimeError(error) + return self + + def stop(self): + """Stop sampling and return a stable copy of the raw snapshots.""" + self._stop_event.set() + if self._thread is not None: + if self._remote_sampler is not None: + self._thread.join(timeout=self.sample_interval + 2) + self._stop_remote_sampler() + self._thread.join(timeout=5) + else: + self._thread.join(timeout=5) + try: + self._append_sample(self._snapshot()) + except OSError as error: + self.error = self.error or str(error) + self._thread = None + else: + self._stop_remote_sampler() + with self._lock: + return list(self._samples) + + def write_case_metrics(self, path, windows): + """Stop sampling and write metrics attributed to completed cases.""" + samples = self.stop() + cases = [] + for window in windows: + # The samples of a failed sampler have gaps + metrics = [] if self.error else metrics_for_window( + samples, window["start"], window["end"]) + cases.append({**window, "metrics": metrics}) + data = {"version": 1, "cases": cases} + if self.error: + data["error"] = self.error + temporary = path + ".tmp" + try: + with open(temporary, "w", encoding="utf-8") as fp: + json.dump(data, fp, separators=(",", ":")) + fp.flush() + os.fsync(fp.fileno()) + os.replace(temporary, path) + except (OSError, TypeError, ValueError): + try: + os.unlink(temporary) + except OSError: + pass + raise + return data + + +def _interpolate_number(before, after, ratio): + return before + (after - before) * ratio + + +def _interpolate_host(before, after, ratio): + cpus = {} + for cpu in before["cpu"].keys() & after["cpu"].keys(): + cpus[cpu] = { + field: _interpolate_number(before["cpu"][cpu][field], + after["cpu"][cpu][field], ratio) + for field in CPU_FIELDS + } + nic = { + name: _interpolate_number(before["nic"][name], after["nic"][name], + ratio) + for name in NIC_METRICS + } + return {"cpu": cpus, "nic": nic} + + +def _snapshot_at(samples, timestamp): + if timestamp <= samples[0]["time"]: + return {"time": timestamp, "hosts": samples[0]["hosts"]} + if timestamp >= samples[-1]["time"]: + return {"time": timestamp, "hosts": samples[-1]["hosts"]} + + for before, after in zip(samples, samples[1:]): + if before["time"] <= timestamp <= after["time"]: + elapsed = after["time"] - before["time"] + ratio = (timestamp - before["time"]) / elapsed + hosts = {} + for host in before["hosts"].keys() & after["hosts"].keys(): + hosts[host] = _interpolate_host( + before["hosts"][host], after["hosts"][host], ratio) + return {"time": timestamp, "hosts": hosts} + raise ValueError(f"no samples around timestamp {timestamp}") + + +def _window_samples(samples, start, end): + # Only the samples tell what happened, a sampler which stopped early must + # not pass for a quiet host + if not samples or end <= start or samples[0]["time"] > start or \ + samples[-1]["time"] < end: + return [] + return [_snapshot_at(samples, start)] + [ + sample for sample in samples if start < sample["time"] < end + ] + [_snapshot_at(samples, end)] + + +def _cpu_deltas(before, after): + result = {} + for cpu in before.keys() & after.keys(): + delta = { + field: after[cpu][field] - before[cpu][field] + for field in CPU_FIELDS + } + if all(value >= 0 for value in delta.values()): + result[cpu] = delta + return result + + +def _per_cpu_array(values): + if not values: + return None + result = [None] * (max(values) + 1) + for cpu, value in values.items(): + result[cpu] = value + return result + + +def _metric(name, metric_type, timestamp, value, host, **policy): + return { + "name": name, + "type": metric_type, + "time": round(timestamp, 6), + "value": value, + "host": host, + "origin": "system-monitor", + **policy, + } + + +def metrics_for_window(samples, start, end): + """Convert raw cumulative snapshots into one testcase's metric series.""" + selected = _window_samples(samples, start, end) + if len(selected) < 2: + return [] + + metrics = [] + hosts = set.intersection(*(set(sample["hosts"]) for sample in selected)) + for host in sorted(hosts): + baseline = selected[0]["hosts"][host] + previous = baseline + for index, sample in enumerate(selected): + current = sample["hosts"][host] + timestamp = sample["time"] - start + elapsed = _cpu_deltas(baseline["cpu"], current["cpu"]) + for name, fields in CPU_TIME_METRICS.items(): + values = _per_cpu_array({ + cpu: sum(delta[field] for field in fields) / CLOCK_TICKS + for cpu, delta in elapsed.items() + }) + if values is not None: + metrics.append(_metric( + name, "per-cpu", timestamp, values, host, + summarize="p90", transform="rate", + aggregation="busy-core-equivalents", + display_range={"min": 0, "max": 100}, + display_scale=100, + display_label="Percent of one CPU")) + + for counter, name in NIC_METRICS.items(): + metrics.append(_metric( + name, "scalar", timestamp, + current["nic"][counter] - baseline["nic"][counter], host, + summarize="max", transform="rate", + display_range={"min": 0})) + + if index: + interval = _cpu_deltas(previous["cpu"], current["cpu"]) + utilization = {} + for cpu, delta in interval.items(): + total = sum(delta.values()) + if total > 0: + busy = total - delta["idle"] - delta["iowait"] + utilization[cpu] = round(100 * busy / total, 4) + values = _per_cpu_array(utilization) + if values is not None: + metrics.append(_metric( + "cpu.utilization", "per-cpu", timestamp, values, + host, summarize="p90", + aggregation="busy-core-equivalents", + display_range={"min": 0, "max": 100}, + display_label="Percent of one CPU")) + previous = current + return metrics diff --git a/contest/hw/tests/test_system_monitor.py b/contest/hw/tests/test_system_monitor.py new file mode 100644 index 00000000..3055f94c --- /dev/null +++ b/contest/hw/tests/test_system_monitor.py @@ -0,0 +1,105 @@ +# SPDX-License-Identifier: GPL-2.0 + +import json +import os +import tempfile +import unittest +from unittest import mock + +from contest.hw.lib.system_monitor import SystemMonitor, metrics_for_window + + +def _cpu(user, idle): + return { + 0: { + 'user': float(user), 'nice': 0.0, 'sys': 0.0, + 'idle': float(idle), 'iowait': 0.0, 'irq': 0.0, + 'sirq': 0.0, 'steal': 0.0, + }, + } + + +def _sample(timestamp, user, idle, dropped): + return { + 'time': timestamp, + 'hosts': { + 'local': { + 'cpu': _cpu(user, idle), + 'nic': { + 'rx_dropped': float(dropped), + 'rx_missed_errors': 0.0, + 'tx_dropped': 0.0, + }, + }, + }, + } + + +class TestMetricWindows(unittest.TestCase): + def test_interpolates_window_boundaries(self): + metrics = metrics_for_window([ + _sample(0.0, 0, 0, 0), + _sample(1.0, 50, 50, 10), + _sample(2.0, 100, 100, 20), + ], 0.5, 1.5) + + user = [metric for metric in metrics + if metric['name'] == 'cpu.time.usr'] + self.assertEqual([metric['time'] for metric in user], [0.0, 0.5, 1.0]) + self.assertEqual([metric['value'] for metric in user], + [[0.0], [0.25], [0.5]]) + + utilization = [metric for metric in metrics + if metric['name'] == 'cpu.utilization'] + self.assertEqual([metric['value'] for metric in utilization], + [[50.0], [50.0]]) + + dropped = [metric for metric in metrics + if metric['name'] == 'nic.rx.dropped'] + self.assertEqual([metric['value'] for metric in dropped], + [0.0, 5.0, 10.0]) + + def test_windows_the_samples_miss_have_no_metrics(self): + samples = [_sample(1.0, 0, 0, 0), _sample(2.0, 50, 50, 10)] + + # Sampling stopped before the end, or started after the beginning + self.assertEqual(metrics_for_window(samples, 1.5, 3.0), []) + self.assertEqual(metrics_for_window(samples, 0.5, 1.5), []) + self.assertNotEqual(metrics_for_window(samples, 1.0, 2.0), []) + + +class TestSystemMetricWriter(unittest.TestCase): + def test_failed_sampler_writes_no_metrics(self): + with tempfile.TemporaryDirectory() as temporary: + path = os.path.join(temporary, 'system-metrics.json') + monitor = mock.Mock(error='remote system sampler stopped') + monitor.stop.return_value = [ + _sample(0.0, 0, 0, 0), _sample(2.0, 100, 100, 20), + ] + + data = SystemMonitor.write_case_metrics( + monitor, path, [{'name': 'case', 'start': 0.5, 'end': 1.5}]) + + self.assertEqual(data['cases'], [{ + 'name': 'case', 'start': 0.5, 'end': 1.5, 'metrics': [], + }]) + self.assertEqual(data['error'], 'remote system sampler stopped') + + def test_writes_sidecar_atomically(self): + with tempfile.TemporaryDirectory() as temporary: + path = os.path.join(temporary, 'system-metrics.json') + monitor = mock.Mock(error=None) + monitor.stop.return_value = [] + + with mock.patch('contest.hw.lib.system_monitor.os.replace', + wraps=os.replace) as replace: + data = SystemMonitor.write_case_metrics(monitor, path, []) + + replace.assert_called_once_with(path + '.tmp', path) + self.assertFalse(os.path.exists(path + '.tmp')) + with open(path, encoding='utf-8') as fp: + self.assertEqual(json.load(fp), data) + + +if __name__ == '__main__': + unittest.main() From 0f7455ac29d9fa83d7900242684cee67bfa890f8 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:47:54 -0700 Subject: [PATCH 14/18] contest: hw: attach system metrics to test results Merge per-case system monitor sidecars into parsed KTAP results. Identify whether each metric came from the selftest or monitor. Signed-off-by: Stanislav Fomichev --- contest/hw/lib/result_parser.py | 53 ++++++++++++++++++ contest/hw/tests/test_system_monitor.py | 71 +++++++++++++++++++++++++ 2 files changed, 124 insertions(+) diff --git a/contest/hw/lib/result_parser.py b/contest/hw/lib/result_parser.py index e9484a82..9ba68b89 100644 --- a/contest/hw/lib/result_parser.py +++ b/contest/hw/lib/result_parser.py @@ -4,11 +4,56 @@ import json import os +from collections import defaultdict, deque from .nipa import guess_indicators, namify, parse_nested_tests, \ result_from_indicators +def _system_metrics(path): + if not os.path.exists(path): + return {} + + try: + with open(path, encoding='utf-8') as fp: + data = json.load(fp) + if not isinstance(data, dict) or data.get('version') != 1 or \ + not isinstance(data.get('cases'), list): + raise ValueError('unsupported system metrics format') + + cases = defaultdict(deque) + for case in data['cases']: + if not isinstance(case, dict) or \ + not isinstance(case.get('name'), str) or \ + not isinstance(case.get('metrics'), list) or \ + not all(isinstance(metric, dict) + for metric in case['metrics']): + raise ValueError('invalid system metrics case') + for metric in case['metrics']: + metric.setdefault('origin', 'system-monitor') + cases[namify(case['name'])].append(case['metrics']) + return cases + except (OSError, TypeError, ValueError) as error: + print(f'parse_results: ignoring {path}: {error}') + return {} + + +def _set_metric_origin(nested): + for case in nested: + for field in ('metrics', 'retry_metrics'): + for metric in case.get(field, []): + metric.setdefault('origin', 'selftest') + + +def _merge_system_metrics(nested, path, retry=False): + case_metrics = _system_metrics(path) + field = 'retry_metrics' if retry else 'metrics' + for case in nested: + matches = case_metrics.get(case['test']) + if matches: + case.setdefault(field, []).extend(matches.popleft()) + + def parse_results(results_path, link): """Parse fetched test output into a vmksft-p-style result list. @@ -77,6 +122,9 @@ def parse_results(results_path, link): # Parse nested subtests nested = parse_nested_tests(stdout, namify) + _set_metric_origin(nested) + _merge_system_metrics( + nested, os.path.join(test_dir, 'system-metrics.json')) # Determine retry result if present retry_result = None @@ -95,6 +143,11 @@ def parse_results(results_path, link): retry_nested = list(nested) parse_nested_tests(retry_stdout, namify, prev_results=retry_nested) + _set_metric_origin(retry_nested or []) + _merge_system_metrics( + retry_nested or [], + os.path.join(retry_dir, 'system-metrics.json'), + retry=True) safe_name = namify(prog) diff --git a/contest/hw/tests/test_system_monitor.py b/contest/hw/tests/test_system_monitor.py index 3055f94c..0cfeb18d 100644 --- a/contest/hw/tests/test_system_monitor.py +++ b/contest/hw/tests/test_system_monitor.py @@ -4,8 +4,10 @@ import os import tempfile import unittest +from pathlib import Path from unittest import mock +from contest.hw.lib.result_parser import parse_results from contest.hw.lib.system_monitor import SystemMonitor, metrics_for_window @@ -101,5 +103,74 @@ def test_writes_sidecar_atomically(self): self.assertEqual(json.load(fp), data) +class TestSystemMetricMerge(unittest.TestCase): + def test_merges_sidecar_by_case_name(self): + with tempfile.TemporaryDirectory() as temporary: + root = Path(temporary) + output = root / 'test-outputs' / '0-example-py' + output.mkdir(parents=True) + (output / 'info').write_text(json.dumps({ + 'retcode': 0, + 'target': 'drivers/net/hw', + 'prog': 'example.py', + }), encoding='utf-8') + (output / 'stdout').write_text( + 'TAP version 13\n' + '1..1\n' + '# TAP version 13\n' + '# 1..1\n' + '# ok 1 first case\n' + 'ok 1 selftests: drivers/net/hw: example.py\n', + encoding='utf-8') + metric = { + 'name': 'cpu.utilization', + 'type': 'per-cpu', + 'time': 1.0, + 'value': [50.0], + 'host': 'local', + 'summarize': 'p90', + 'aggregation': 'busy-core-equivalents', + } + (output / 'system-metrics.json').write_text(json.dumps({ + 'version': 1, + 'cases': [{'name': 'first case', 'metrics': [metric]}], + }), encoding='utf-8') + + results = parse_results(str(root), '.') + metric['origin'] = 'system-monitor' + self.assertEqual(results[0]['results'][0]['metrics'], [metric]) + + def test_ignores_invalid_sidecar(self): + with tempfile.TemporaryDirectory() as temporary: + root = Path(temporary) + output = root / 'test-outputs' / '0-example-py' + output.mkdir(parents=True) + (output / 'info').write_text(json.dumps({ + 'retcode': 0, + 'target': 'drivers/net/hw', + 'prog': 'example.py', + }), encoding='utf-8') + (output / 'stdout').write_text( + '# TAP version 13\n# 1..1\n# ok 1 first case\n', + encoding='utf-8') + sidecar = output / 'system-metrics.json' + + invalid_data = ( + '{', + json.dumps({ + 'version': 1, + 'cases': [{'name': 'first case'}], + }), + ) + for data in invalid_data: + with self.subTest(data=data): + sidecar.write_text(data, encoding='utf-8') + with mock.patch('builtins.print') as warning: + results = parse_results(str(root), '.') + warning.assert_called_once() + self.assertNotIn( + 'metrics', results[0]['results'][0]) + + if __name__ == '__main__': unittest.main() From 5409e5bb99590415f11d38842aae468b643049ca Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:48:07 -0700 Subject: [PATCH 15/18] contest: hw: collect system metrics of performance test cases Sample the DUT and peer around each performance test case when given a system monitor configuration. Merge the samples of the cases of a program into its system-metrics.json, keeping the error of a failed sampler with the cases it left without metrics. Signed-off-by: Stanislav Fomichev --- contest/hw/lib/runner.py | 88 +++++++++++++++++----- contest/hw/tests/test_hw_worker.py | 115 ++++++++++++++++++++++++++++- 2 files changed, 184 insertions(+), 19 deletions(-) diff --git a/contest/hw/lib/runner.py b/contest/hw/lib/runner.py index 1236bfd9..c61d3f18 100644 --- a/contest/hw/lib/runner.py +++ b/contest/hw/lib/runner.py @@ -12,6 +12,7 @@ from dataclasses import dataclass from lib.nipa import has_crash, extract_crash, namify, parse_nested_tests +from lib.system_monitor import SystemMonitor # Default per-test wall-clock limit before we start tearing the test down @@ -305,7 +306,7 @@ def _list_subtests(test_dir, target, prog, timeout): def _run_one_test(test_dir, output_dir, target, prog, timeout, - case_name=None): + monitor_config=None, case_name=None): """Run a single test and save stdout/stderr. Returns (retcode, elapsed). On timeout the test is asked to shut down gracefully (SIGINT, which @@ -322,6 +323,16 @@ def _run_one_test(test_dir, output_dir, target, prog, timeout, is reached -- otherwise it is orphaned and lingers, poisoning the retry and later crashing on its own. """ + monitor = None + case_start = None + if monitor_config: + try: + monitor = SystemMonitor(**monitor_config).start() + case_start = monitor.elapsed() + except Exception as error: # pylint: disable=broad-except + print(f" {target}:{prog}: system monitor unavailable: {error}") + monitor = None + env = os.environ.copy() if case_name is not None: args_name = _test_args_name(prog) @@ -329,14 +340,19 @@ def _run_one_test(test_dir, output_dir, target, prog, timeout, t1 = time.monotonic() # pylint: disable=consider-using-with - proc = subprocess.Popen( - ['./run_kselftest.sh', '-t', f'{target}:{prog}'], - cwd=test_dir, - env=env, - stdout=subprocess.PIPE, - stderr=subprocess.PIPE, - start_new_session=True, - ) + try: + proc = subprocess.Popen( + ['./run_kselftest.sh', '-t', f'{target}:{prog}'], + cwd=test_dir, + env=env, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE, + start_new_session=True, + ) + except Exception: + if monitor is not None: + monitor.stop() + raise # With start_new_session the child is its own group leader (pgid == pid). pgid = proc.pid stop_kind = None @@ -363,8 +379,22 @@ def _run_one_test(test_dir, output_dir, target, prog, timeout, _kill_session(proc.pid, signal.SIGKILL) out, err = proc.communicate() + t2 = time.monotonic() stdout = (out or b'').decode('utf-8', 'ignore') stderr = (err or b'').decode('utf-8', 'ignore') + if monitor is not None: + try: + monitor.write_case_metrics( + os.path.join(output_dir, 'system-metrics.json'), [{ + 'name': case_name, + 'start': case_start, + 'end': monitor.elapsed(), + }]) + except Exception as error: # pylint: disable=broad-except + print(f" {target}:{prog}: failed to save system metrics: {error}") + if monitor.error: + print(f" {target}:{prog}: system monitor failed: {monitor.error}") + if stop_kind is not None: # Preserve whatever the test printed, but mark the timeout in # both streams so it's visible whichever one is inspected. @@ -373,7 +403,6 @@ def _run_one_test(test_dir, output_dir, target, prog, timeout, stderr += marker elif not stdout and not stderr: print(f" {target}:{prog}: no output (rc={retcode})") - t2 = time.monotonic() elapsed = round(t2 - t1, 1) with open(os.path.join(output_dir, 'stdout'), 'w', encoding='utf-8') as fp: @@ -392,7 +421,8 @@ def _read_output(path, name): return output -def _run_performance_test(test_dir, output_dir, target, prog, timeout): +def _run_performance_test(test_dir, output_dir, target, prog, timeout, + monitor_config=None): """Run each case in a performance test program in its own process.""" try: cases = _list_subtests(test_dir, target, prog, timeout) @@ -411,6 +441,7 @@ def _run_performance_test(test_dir, output_dir, target, prog, timeout): stderr = [] retcodes = [] elapsed = 0.0 + system_cases = [] try: for index, case_name in enumerate(cases): @@ -419,12 +450,24 @@ def _run_performance_test(test_dir, output_dir, target, prog, timeout): os.makedirs(case_dir) retcode, case_elapsed = _run_one_test( test_dir, case_dir, target, prog, timeout, - case_name=case_name) + monitor_config=monitor_config, case_name=case_name) retcodes.append(retcode) elapsed += case_elapsed stdout.append(_read_output(case_dir, 'stdout')) stderr.append(_read_output(case_dir, 'stderr')) + metrics_path = os.path.join(case_dir, 'system-metrics.json') + if os.path.exists(metrics_path): + with open(metrics_path, encoding='utf-8') as fp: + case_metrics = json.load(fp) + module = os.path.splitext(os.path.basename(prog))[0] + for case in case_metrics['cases']: + case['name'] = f"{module}.{case['name']}" + # Keep why the sampler left the case without metrics + if case_metrics.get('error'): + case['error'] = case_metrics['error'] + system_cases.extend(case_metrics['cases']) + if _hit_timeout(case_dir): break finally: @@ -434,6 +477,11 @@ def _run_performance_test(test_dir, output_dir, target, prog, timeout): fp.write(''.join(stdout)) with open(os.path.join(output_dir, 'stderr'), 'w', encoding='utf-8') as fp: fp.write(''.join(stderr)) + if system_cases: + with open(os.path.join(output_dir, 'system-metrics.json'), 'w', + encoding='utf-8') as fp: + json.dump({'version': 1, 'cases': system_cases}, fp, + separators=(',', ':')) if any(retcode not in (0, 4) for retcode in retcodes): retcode = next(retcode for retcode in retcodes @@ -445,15 +493,17 @@ def _run_performance_test(test_dir, output_dir, target, prog, timeout): return retcode, round(elapsed, 1) -def _execute_test(test_dir, output_dir, target, prog, timeout): +def _execute_test(test_dir, output_dir, target, prog, timeout, + monitor_config=None): if target == PERFORMANCE_TARGET or \ target.startswith(PERFORMANCE_TARGET + '/'): return _run_performance_test( - test_dir, output_dir, target, prog, timeout) + test_dir, output_dir, target, prog, timeout, monitor_config) return _run_one_test(test_dir, output_dir, target, prog, timeout) -def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): +def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT, + monitor_config=None): """Execute kselftest in 'installed' form. For each test: @@ -461,7 +511,7 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): 2. Check if test_name is in .attempted -- if so, skip (crash recovery) 3. Write test_name to .attempted + fsync (crash-safe bookkeeping) 4. Run via: ./run_kselftest.sh -t :. Performance - programs run once per listed subtest. + programs run once per listed subtest with system monitoring. 5. Capture stdout/stderr, save to results_dir// 6. Drain dmesg output produced during the test, save to dmesg file 7. Save metadata (retcode, time, target, prog) to info file @@ -524,7 +574,8 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): # Run the test retcode, elapsed = _execute_test( - test_dir, test_results_dir, target, prog, timeout) + test_dir, test_results_dir, target, prog, timeout, + monitor_config) with open(os.path.join(test_results_dir, 'stdout'), encoding='utf-8') as fp: stdout = fp.read() @@ -565,7 +616,8 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT): retry_dir = os.path.join(results_dir, f'{dir_name}-retry') os.makedirs(retry_dir, exist_ok=True) retry_retcode, _retry_elapsed = _execute_test( - test_dir, retry_dir, target, prog, timeout) + test_dir, retry_dir, target, prog, timeout, + monitor_config) if _hit_timeout(retry_dir): tainted = True print(f"[{test_idx+1}/{len(tests)}] {test_name}: retry hit " diff --git a/contest/hw/tests/test_hw_worker.py b/contest/hw/tests/test_hw_worker.py index ba8c7c9d..226fc361 100644 --- a/contest/hw/tests/test_hw_worker.py +++ b/contest/hw/tests/test_hw_worker.py @@ -1177,7 +1177,7 @@ def test_performance_program_runs_each_subtest_once(self): calls = [] def run_case(_test_dir, case_dir, _target, _prog, _timeout, - case_name=None): + monitor_config=None, case_name=None): calls.append(case_name) Path(case_dir, 'stdout').write_text( '# TAP version 13\n# 1..1\n' @@ -1218,5 +1218,118 @@ def test_performance_target_runs_cases_one_at_a_time(self): regular.assert_called_once_with( temporary, temporary, 'drivers/net/hw', 'regular.py', 10) + +class FakeMonitor: + instances = [] + + def __init__(self, **_kwargs): + self.windows = None + self.elapsed_time = 0.0 + self.error = None + self.__class__.instances.append(self) + + def start(self): + return self + + def elapsed(self): + self.elapsed_time += 1.0 + return self.elapsed_time + + def write_case_metrics(self, path, windows): + self.windows = windows + Path(path).write_text(json.dumps({ + 'version': 1, + 'cases': [{**window, 'metrics': []} for window in windows], + }), encoding='utf-8') + + +class TestMonitoredExecution(unittest.TestCase): + def test_uses_process_boundaries_for_one_subtest(self): + FakeMonitor.instances.clear() + with tempfile.TemporaryDirectory() as temporary: + test_dir = Path(temporary) / 'tests' + output_dir = Path(temporary) / 'output' + test_dir.mkdir() + output_dir.mkdir() + + with mock.patch.object(runner, 'SystemMonitor', FakeMonitor), \ + mock.patch.object(runner.subprocess, 'Popen', + return_value=FakeProcess()) as popen: + retcode, _elapsed = runner._run_one_test( + str(test_dir), str(output_dir), + runner.PERFORMANCE_TARGET, 'example.py', 10, + monitor_config={'ifname': 'eth0'}, + case_name='measured.case') + + self.assertEqual(retcode, 0) + self.assertEqual(FakeMonitor.instances[0].windows, [{ + 'name': 'measured.case', 'start': 1.0, 'end': 2.0, + }]) + environment = popen.call_args.kwargs['env'] + self.assertEqual(environment['KSELFTEST_EXAMPLE_PY_ARGS'], + '-t measured.case') + self.assertTrue((output_dir / 'system-metrics.json').exists()) + + def test_merges_system_metrics_of_each_subtest(self): + with tempfile.TemporaryDirectory() as temporary: + test_dir = Path(temporary) / 'tests' + output_dir = Path(temporary) / 'output' + test_dir.mkdir() + output_dir.mkdir() + calls = [] + + def run_case(_test_dir, case_dir, _target, _prog, _timeout, + monitor_config=None, case_name=None): + calls.append((case_name, monitor_config)) + Path(case_dir, 'stdout').write_text('', encoding='utf-8') + Path(case_dir, 'stderr').write_text('', encoding='utf-8') + sidecar = { + 'version': 1, + 'cases': [{'name': case_name, 'metrics': []}], + } + if case_name == 'case.two': + sidecar['error'] = 'sampler stopped' + Path(case_dir, 'system-metrics.json').write_text( + json.dumps(sidecar), encoding='utf-8') + return 0, 1.0 + + with mock.patch.object(runner, '_list_subtests', + return_value=['case.one', 'case.two']), \ + mock.patch.object(runner, '_run_one_test', + side_effect=run_case): + runner._run_performance_test( + str(test_dir), str(output_dir), + runner.PERFORMANCE_TARGET, 'example.py', 10, + {'ifname': 'eth0'}) + + self.assertEqual(calls, [ + ('case.one', {'ifname': 'eth0'}), + ('case.two', {'ifname': 'eth0'}), + ]) + with open(Path(output_dir, 'system-metrics.json'), + encoding='utf-8') as fp: + metrics = json.load(fp) + self.assertEqual([case['name'] for case in metrics['cases']], + ['example.case.one', 'example.case.two']) + self.assertEqual([case.get('error') for case in metrics['cases']], + [None, 'sampler stopped']) + + def test_monitoring_is_selected_by_target(self): + with tempfile.TemporaryDirectory() as temporary: + with mock.patch.object(runner, '_run_performance_test', + return_value=(0, 1.0)) as performance, \ + mock.patch.object(runner, '_run_one_test', + return_value=(0, 1.0)) as regular: + runner._execute_test( + temporary, temporary, runner.PERFORMANCE_TARGET, + 'perf.py', 10, {'ifname': 'eth0'}) + runner._execute_test( + temporary, temporary, 'drivers/net/hw', + 'regular.py', 10, {'ifname': 'eth0'}) + + performance.assert_called_once() + regular.assert_called_once_with( + temporary, temporary, 'drivers/net/hw', 'regular.py', 10) + if __name__ == '__main__': unittest.main() From 46e5ee4207a8bcba2461207b91eed18066210000 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:48:24 -0700 Subject: [PATCH 16/18] contest: hw: run installed selftests directly Add a worker mode which runs an installed selftest tree with existing host networking. Generate net.config files from the interface addresses, unless nic-test.env gives them, which it has to for LOCAL_PREFIX_V6. The kernel was not booted for such a run, so ignore what it logged before. Share the setup with hardware runners, which now also pass the system monitor configuration to run_tests(), so their performance tests collect system metrics, and export the non-empty KSELFTEST_* variables of nic-test.env, like per-program arguments. Signed-off-by: Stanislav Fomichev --- contest/hw/hw_worker.py | 162 +++++++++++++++++++++++++---- contest/hw/lib/runner.py | 21 ++-- contest/hw/tests/test_hw_worker.py | 134 ++++++++++++++++++++++++ 3 files changed, 288 insertions(+), 29 deletions(-) diff --git a/contest/hw/hw_worker.py b/contest/hw/hw_worker.py index b52f7a3d..105535d4 100644 --- a/contest/hw/hw_worker.py +++ b/contest/hw/hw_worker.py @@ -69,6 +69,21 @@ def _parse_env_file(path): return env +def _system_monitor_config(env): + """Return system monitor arguments for performance tests.""" + if not env.get('NETIF'): + return None + + config = {'ifname': env['NETIF']} + if env.get('REMOTE_TYPE') == 'ssh' and env.get('REMOTE_ARGS') and \ + env.get('REMOTE_IFNAME'): + config.update({ + 'remote': env['REMOTE_ARGS'], + 'remote_ifname': env['REMOTE_IFNAME'], + }) + return config + + def _test_timeout(env): """Per-test timeout from nic-test.env, or the runner default. @@ -406,6 +421,82 @@ def _setup_irq_affinity(netif): print(f"WARN: could not restore combined channels on {netif}") +def _interface_addresses(ifname, host=None): + """Return existing global IPv4 and IPv6 addresses on an interface.""" + result = _ip(f'-json address show dev {ifname}', host=host) + try: + links = json.loads(result.stdout) + except json.JSONDecodeError as error: + raise RuntimeError(f'failed to read addresses for {ifname}') from error + + addresses = {} + for address in links[0]['addr_info']: + family = address['family'] + if address['scope'] == 'global' and family in ('inet', 'inet6') \ + and family not in addresses: + addresses[family] = address['local'] + return addresses + + +def _selected_targets(test_dir): + path = os.path.join(test_dir, 'kselftest-list.txt') + with open(path, encoding='utf-8') as fp: + return {line.strip().partition(':')[0] for line in fp} + + +def _write_discovered_net_config(test_dir, env): + """Describe existing DUT and peer networking without changing it.""" + targets = [] + for target in _selected_targets(test_dir): + target_dir = os.path.join(test_dir, target) + config_path = os.path.join(target_dir, 'net.config') + if os.path.isdir(target_dir) and not os.path.exists(config_path): + targets.append((target, config_path)) + if not targets: + return + + ifname = env.get('NETIF') + remote_ifname = env.get('REMOTE_IFNAME') + remote = env.get('REMOTE_ARGS') + if not ifname or not remote_ifname or not remote: + raise RuntimeError('local mode requires DUT and peer interfaces') + + local = _interface_addresses(ifname) + peer = _interface_addresses(remote_ifname, host=remote) + config = { + 'NETIF': ifname, + 'REMOTE_TYPE': 'ssh', + 'REMOTE_ARGS': remote, + } + if 'inet' in local and 'inet' in peer: + config['LOCAL_V4'] = local['inet'] + config['REMOTE_V4'] = peer['inet'] + if 'inet6' in local and 'inet6' in peer: + config['LOCAL_V6'] = local['inet6'] + config['REMOTE_V6'] = peer['inet6'] + # Given values win, and a prefix routed to the DUT, LOCAL_PREFIX_V6, can + # only be given + for key in _NET_CONFIG_KEYS: + if env.get(key): + config[key] = env[key] + if 'LOCAL_V4' not in config and 'LOCAL_V6' not in config: + raise RuntimeError('DUT and peer have no common global address family') + + content = ''.join(f'{key}={value}\n' for key, value in config.items()) + for _target, path in targets: + with open(path, 'w', encoding='utf-8') as fp: + fp.write(content) + print(f"Wrote {path} from existing interface addresses") + + +def _export_test_env(env): + """Export runner-independent kselftest environment variables.""" + for key, value in env.items(): + # ksft cannot parse an empty DISRUPTIVE, leave unset what is empty + if value and (key in _ENV_ONLY_KEYS or key.startswith('KSELFTEST_')): + os.environ[key] = value + + def setup_test_interfaces(test_dir): """Configure test NICs and write net.config from nic-test.env. @@ -496,10 +587,7 @@ def setup_test_interfaces(test_dir): fp.write(config_content) print(f"Wrote {path}") - # Export env-only variables (not net.config) for the test framework - for key in _ENV_ONLY_KEYS: - if env.get(key): - os.environ[key] = env[key] + _export_test_env(env) def _find_initrd(): @@ -537,6 +625,33 @@ def _kexec_local(bzimage): subprocess.run(['kexec', '-e'], check=False) +def _run_test_dir(test_dir, results_dir, booted=True): + """Configure and execute one installed kselftest directory.""" + env = _parse_env_file(os.path.join(test_dir, 'nic-test.env')) + if env.get('NIPA_SKIP_NET_SETUP') == 'yes': + print("Using the existing host network configuration") + _write_discovered_net_config(test_dir, env) + _export_test_env(env) + else: + setup_test_interfaces(test_dir) + os.makedirs(results_dir, exist_ok=True) + + netif = env.get('NETIF') + if netif: + dev_info = _collect_device_info(netif) + if dev_info: + with open(os.path.join(results_dir, 'device-info.json'), 'w', + encoding='utf-8') as fp: + json.dump(dev_info, fp) + print(f"Collected device info for {netif}") + else: + print(f"Warning: could not collect device info for {netif}") + + return run_tests( + test_dir, results_dir, timeout=_test_timeout(env), + monitor_config=_system_monitor_config(env), booted=booted) + + def main(): """Find pending tests, run them, and write results.""" sys.stdout = _TeeWriter(sys.stdout) @@ -554,6 +669,10 @@ def main(): # on-DUT equivalent of deployer.kexec_machine(). group.add_argument('--kexec-only', metavar='DIR', help="load DIR/bzImage and boot into it, then exit") + group.add_argument('--run-only', metavar='DIR', + help="run an installed kselftest directory directly") + parser.add_argument('--results-dir', metavar='DIR', + help="result directory for --run-only") args = parser.parse_args() if args.prep_only: @@ -573,6 +692,21 @@ def main(): _kexec_local(bzimage) return + if args.run_only: + test_dir = os.path.abspath(args.run_only) + if not os.path.isfile(os.path.join(test_dir, 'run_kselftest.sh')): + print(f"No run_kselftest.sh found in {test_dir}") + sys.exit(1) + results_dir = args.results_dir or os.path.join(test_dir, 'results') + print(test_dir, "Starting tests directly") + outcome = _run_test_dir(test_dir, os.path.abspath(results_dir), + booted=False) + if outcome.crashed: + print("NIPA DETECTED SYSTEM CRASH") + if outcome.tainted: + print("NIPA DETECTED TEST TIMEOUT") + return + tests_dir = TESTS_DIR results_base = RESULTS_DIR @@ -602,27 +736,9 @@ def main(): print(test_dir, "Starting tests") - # Configure test interfaces and write net.config - setup_test_interfaces(test_dir) - reservation_id = os.path.basename(test_dir) results_dir = os.path.join(results_base, reservation_id) - os.makedirs(results_dir, exist_ok=True) - - # Collect devlink device info for the test NIC - env = _parse_env_file(os.path.join(test_dir, 'nic-test.env')) - netif = env.get('NETIF') - if netif: - dev_info = _collect_device_info(netif) - if dev_info: - with open(os.path.join(results_dir, 'device-info.json'), 'w', - encoding='utf-8') as fp: - json.dump(dev_info, fp) - print(f"Collected device info for {netif}") - else: - print(f"Warning: could not collect device info for {netif}") - - outcome = run_tests(test_dir, results_dir, timeout=_test_timeout(env)) + outcome = _run_test_dir(test_dir, results_dir) print(f"Completed, results in {results_dir}") # Sentinels for the orchestrator -- it greps our journal for these and diff --git a/contest/hw/lib/runner.py b/contest/hw/lib/runner.py index c61d3f18..37d3a193 100644 --- a/contest/hw/lib/runner.py +++ b/contest/hw/lib/runner.py @@ -503,9 +503,12 @@ def _execute_test(test_dir, output_dir, target, prog, timeout, def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT, - monitor_config=None): + monitor_config=None, booted=True): """Execute kselftest in 'installed' form. + Without booted, the kernel was not booted for this run, and the messages + it logged before belong to someone else. + For each test: 1. Check dmesg for crash from previous test -- if found, stop 2. Check if test_name is in .attempted -- if so, skip (crash recovery) @@ -536,7 +539,7 @@ def run_tests(test_dir, results_dir, timeout=DEFAULT_TEST_TIMEOUT, known_bad = load_known_bad(test_dir) # Open dmesg once, drain boot messages - dmesg = DmesgReader() + dmesg = DmesgReader(skip_existing=not booted) boot_lines = dmesg.drain() if boot_lines: boot_path = os.path.join(results_dir, 'boot-dmesg') @@ -671,18 +674,24 @@ class DmesgReader: all lines accumulated since the last drain. Not threaded — the caller is expected to drain between tests. - /dev/kmsg gives each opener its own read position. It does not - support seek(), so we just read from wherever the kernel puts us - (typically the start of the ring buffer) and drain forward. + /dev/kmsg gives each opener its own read position, starting at the + oldest record in the ring buffer, and we drain forward from there. + With skip_existing, we start after the newest record instead. """ - def __init__(self): + def __init__(self, skip_existing=False): self._fd = None try: # O_RDONLY | O_NONBLOCK so reads return EAGAIN instead of blocking self._fd = os.open('/dev/kmsg', os.O_RDONLY | os.O_NONBLOCK) except (PermissionError, FileNotFoundError, OSError) as e: print(f"DmesgReader: cannot open /dev/kmsg: {e}") + return + if skip_existing: + try: + os.lseek(self._fd, 0, os.SEEK_END) + except OSError: + self.drain() def drain(self): """Read all available lines from /dev/kmsg. diff --git a/contest/hw/tests/test_hw_worker.py b/contest/hw/tests/test_hw_worker.py index 226fc361..4c2dc8da 100644 --- a/contest/hw/tests/test_hw_worker.py +++ b/contest/hw/tests/test_hw_worker.py @@ -19,6 +19,7 @@ _hit_timeout, _run_one_test, _known_bad_retry_decision) from lib.nipa import namify +import hw_worker from lib import runner @@ -378,6 +379,29 @@ def test_single_test_pass(self, mock_popen, mock_dmesg_cls): info = self._read_info(results_dir) self.assertEqual(info['retcode'], 0) + mock_dmesg_cls.assert_called_once_with(skip_existing=False) + + @mock.patch('lib.runner.DmesgReader') + @mock.patch('subprocess.Popen') + def test_skips_older_kernel_messages_without_boot(self, mock_popen, + mock_dmesg_cls): + mock_dmesg_cls.return_value.drain.return_value = '' + mock_popen.return_value = _fake_popen( + returncode=0, stdout=b'ok 1 test_name\n', stderr=b'') + + with tempfile.TemporaryDirectory() as tmpdir: + test_dir = os.path.join(tmpdir, 'tests') + results_dir = os.path.join(tmpdir, 'results') + os.makedirs(test_dir) + os.makedirs(results_dir) + with open(os.path.join(test_dir, 'kselftest-list.txt'), 'w') as fp: + fp.write('net:test1.sh\n') + + outcome = run_tests(test_dir, results_dir, booted=False) + + mock_dmesg_cls.assert_called_once_with(skip_existing=True) + mock_popen.assert_called_once() + self.assertFalse(outcome.crashed) @mock.patch('lib.runner.DmesgReader') @mock.patch('subprocess.Popen') @@ -850,6 +874,19 @@ def test_close_no_fd(self): dmesg._fd = None dmesg.close() # should not raise + @mock.patch('lib.runner.os.lseek') + @mock.patch('lib.runner.os.open', return_value=1234) + def test_skip_existing_starts_after_newest_record(self, _mock_open, + mock_lseek): + DmesgReader(skip_existing=True) + mock_lseek.assert_called_once_with(1234, 0, os.SEEK_END) + + @mock.patch('lib.runner.os.lseek') + @mock.patch('lib.runner.os.open', return_value=1234) + def test_reads_whole_buffer_by_default(self, _mock_open, mock_lseek): + DmesgReader() + mock_lseek.assert_not_called() + class TestMainFlow(_MainTestCase): def test_no_tests(self): @@ -1331,5 +1368,102 @@ def test_monitoring_is_selected_by_target(self): regular.assert_called_once_with( temporary, temporary, 'drivers/net/hw', 'regular.py', 10) + +class TestDirectWorkerMode(unittest.TestCase): + def test_preserves_existing_network_configuration(self): + with tempfile.TemporaryDirectory() as temporary: + test_dir = Path(temporary) / 'tests' + results_dir = Path(temporary) / 'results' + config = test_dir / 'drivers' / 'net' / 'hw' / 'net.config' + config.parent.mkdir(parents=True) + config.write_text('LOCAL_V6=fd00::1\n', encoding='utf-8') + (test_dir / 'kselftest-list.txt').write_text( + 'drivers/net/hw:example.py\n', encoding='utf-8') + (test_dir / 'nic-test.env').write_text( + 'NIPA_SKIP_NET_SETUP=yes\nNETIF=eth0\n', encoding='utf-8') + + with mock.patch.object(hw_worker, 'setup_test_interfaces') as setup, \ + mock.patch.object(hw_worker, '_collect_device_info', + return_value=None), \ + mock.patch.object(hw_worker, 'run_tests') as run_tests: + hw_worker._run_test_dir(str(test_dir), str(results_dir), + booted=False) + + setup.assert_not_called() + run_tests.assert_called_once() + self.assertIs(run_tests.call_args.kwargs['booted'], False) + self.assertEqual(config.read_text(encoding='utf-8'), + 'LOCAL_V6=fd00::1\n') + + def test_generates_config_from_existing_addresses(self): + with tempfile.TemporaryDirectory() as temporary: + test_dir = Path(temporary) + target = test_dir / 'drivers/net/perf' + target.mkdir(parents=True) + (test_dir / 'kselftest-list.txt').write_text( + 'drivers/net/perf:iperf.py\n', encoding='utf-8') + env = { + 'NETIF': 'eth0', + 'REMOTE_IFNAME': 'eth0', + 'REMOTE_ARGS': 'root@peer', + } + + with mock.patch.object(hw_worker, '_interface_addresses', + side_effect=[ + {'inet': '192.0.2.1', + 'inet6': 'fd00:1::1'}, + {'inet': '192.0.2.2', + 'inet6': 'fd00:1::2'}, + ]) as addresses: + hw_worker._write_discovered_net_config(str(test_dir), env) + + self.assertEqual(addresses.call_args_list, [ + mock.call('eth0'), mock.call('eth0', host='root@peer'), + ]) + config = (target / 'net.config').read_text(encoding='utf-8') + self.assertIn('NETIF=eth0\n', config) + self.assertIn('LOCAL_V4=192.0.2.1\n', config) + self.assertIn('REMOTE_V4=192.0.2.2\n', config) + self.assertIn('LOCAL_V6=fd00:1::1\n', config) + self.assertIn('REMOTE_V6=fd00:1::2\n', config) + self.assertNotIn('LOCAL_PREFIX_V6', config) + self.assertIn('REMOTE_ARGS=root@peer\n', config) + + def test_given_values_replace_discovered_ones(self): + with tempfile.TemporaryDirectory() as temporary: + test_dir = Path(temporary) + target = test_dir / 'drivers/net/perf' + target.mkdir(parents=True) + (test_dir / 'kselftest-list.txt').write_text( + 'drivers/net/perf:iperf.py\n', encoding='utf-8') + env = { + 'NETIF': 'eth0', + 'REMOTE_IFNAME': 'eth0', + 'REMOTE_ARGS': 'root@peer', + 'LOCAL_V6': '2001:db8:1::1', + 'LOCAL_PREFIX_V6': '2001:db8:2::/64', + } + + with mock.patch.object(hw_worker, '_interface_addresses', + side_effect=[{'inet6': 'fd00:1::1'}, + {'inet6': 'fd00:1::2'}]): + hw_worker._write_discovered_net_config(str(test_dir), env) + + config = (target / 'net.config').read_text(encoding='utf-8') + self.assertIn('LOCAL_V6=2001:db8:1::1\n', config) + self.assertIn('REMOTE_V6=fd00:1::2\n', config) + self.assertIn('LOCAL_PREFIX_V6=2001:db8:2::/64\n', config) + + def test_exports_test_variables_with_values(self): + env = {'DISRUPTIVE': '', 'KSFT_MACHINE_SLOW': 'yes', + 'KSELFTEST_IPERF_PY_ARGS': '-t tcp', 'NETIF': 'eth0'} + with mock.patch.dict(hw_worker.os.environ, {}, clear=True): + hw_worker._export_test_env(env) + exported = dict(hw_worker.os.environ) + + self.assertEqual(exported, {'KSFT_MACHINE_SLOW': 'yes', + 'KSELFTEST_IPERF_PY_ARGS': '-t tcp'}) + + if __name__ == '__main__': unittest.main() From 6f019c8c89dc3dfa1f11347a459cee93fabf40e4 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:48:54 -0700 Subject: [PATCH 17/18] contest: hw: add local hardware test runner Stage built selftests on a DUT, prepare routes and peer SSH access, and run selected programs without kernel deployment. Fail, after saving the results, when the worker reports a kernel crash or a test timeout. NIPA does not build or install kperf. Hardware performance runners require: - kperf-server on both hosts - kperf-client on the DUT Signed-off-by: Stanislav Fomichev --- .gitignore | 1 + README.rst | 32 +++ contest/hw/__init__.py | 1 + contest/hw/local.py | 429 +++++++++++++++++++++++++++++++++ contest/hw/tests/test_local.py | 176 ++++++++++++++ nipa | 19 ++ 6 files changed, 658 insertions(+) create mode 100644 contest/hw/__init__.py create mode 100644 contest/hw/local.py create mode 100644 contest/hw/tests/test_local.py create mode 100755 nipa diff --git a/.gitignore b/.gitignore index 572a6a4c..4b459ca1 100644 --- a/.gitignore +++ b/.gitignore @@ -8,3 +8,4 @@ ui/contest/ ui/static/ __pycache__/ results/ +nipa-results/ diff --git a/README.rst b/README.rst index 50e79c59..6e8c6f1d 100644 --- a/README.rst +++ b/README.rst @@ -72,6 +72,38 @@ which takes a long time:: patch/build_clang \ patch/build_clang_rust +Running hardware tests locally +------------------------------ + +How to run +~~~~~~~~~~ + +Requirements: + +* selftests in ``$linux`` are already built; +* root SSH works from the local host to the DUT and peer; +* the selected interfaces have global addresses; +* both hosts provide ``iperf3``, ``/dev/udmabuf``, and ``kperf-server``; +* the DUT also provides ``kperf-client``. + +Run iperf and kperf devmem:: + + cd $nipa + ./nipa local run \ + --tree $linux \ + --dut root@dut.example.com \ + --peer root@peer.example.com \ + --netif eth0 --peer-netif eth0 \ + --test drivers/net/perf + +``net.config`` takes the addresses of the interfaces, pass ``--env KEY=VALUE`` +for others, e.g. ``LOCAL_PREFIX_V6``. The run fails if the worker detected a +kernel crash or a test timeout. + +Results are written to ``$PWD/nipa-results/``. Use +``target:program`` to select one program or ``--output`` to change the results +directory. + Structure ========= diff --git a/contest/hw/__init__.py b/contest/hw/__init__.py new file mode 100644 index 00000000..f66554cd --- /dev/null +++ b/contest/hw/__init__.py @@ -0,0 +1 @@ +# SPDX-License-Identifier: GPL-2.0 diff --git a/contest/hw/local.py b/contest/hw/local.py new file mode 100644 index 00000000..ce5ed410 --- /dev/null +++ b/contest/hw/local.py @@ -0,0 +1,429 @@ +# SPDX-License-Identifier: GPL-2.0 + +"""Database-free local frontend for the NIPA hardware worker.""" + +import argparse +import datetime +import json +import os +from pathlib import Path +import shlex +import shutil +import subprocess +import sys +import tempfile +import uuid + +from contest.hw.lib.result_parser import parse_results + + +HW_DIR = Path(__file__).resolve().parent +ROOT_DIR = HW_DIR.parent.parent + + +SSH_OPTIONS = [ + '-o', 'ConnectTimeout=10', + '-o', 'StrictHostKeyChecking=no', + '-o', 'BatchMode=yes', +] + + +def _run(command, **kwargs): + print('+', shlex.join(str(part) for part in command)) + kwargs.setdefault('check', True) + return subprocess.run(command, **kwargs) + + +def _ssh(host, command, **kwargs): + return _run(['ssh', *SSH_OPTIONS, host, command], **kwargs) + + +def _rsync(source, destination): + ssh_command = shlex.join(['ssh', *SSH_OPTIONS]) + _run(['rsync', '-a', '--exclude=__pycache__', '-e', ssh_command, + str(source), str(destination)]) + + +def _ssh_json(host, command): + completed = _ssh(host, shlex.join(command), capture_output=True, text=True) + try: + return json.loads(completed.stdout) + except json.JSONDecodeError as error: + raise RuntimeError(f'invalid JSON from {host}: {command[0]}') from error + + +def _host_addresses(host, ifname): + links = _ssh_json(host, ['ip', '-j', 'address', 'show', 'dev', ifname]) + addresses = {} + for address in links[0]['addr_info']: + family = address['family'] + if address['scope'] == 'global' and family in ('inet', 'inet6') \ + and family not in addresses: + addresses[family] = address['local'] + return addresses + + +def _prepare_route(host, family, source, destination, ifname): + current = _ssh_json( + host, ['ip', '-j', f'-{family}', 'route', 'get', destination])[0] + if current.get('dev') == ifname and current.get('prefsrc') == source: + return None + + desired = _ssh_json(host, [ + 'ip', '-j', f'-{family}', 'route', 'get', destination, + 'from', source, + ])[0] + prefix = f'{destination}/{"32" if family == 4 else "128"}' + route = ['ip', f'-{family}', 'route', 'add', prefix] + if desired.get('gateway'): + route.extend(['via', desired['gateway']]) + route.extend(['dev', ifname, 'src', source, 'proto', 'static']) + _ssh(host, shlex.join(route)) + route[3] = 'del' + print(f'{host}: routing {destination} over {ifname}') + return host, route + + +def _remove_routes(routes): + for host, route in reversed(routes): + completed = _ssh(host, shlex.join(route), check=False) + if completed.returncode: + print(f'Warning: failed to remove temporary route on {host}', + file=sys.stderr) + + +def _prepare_routes(dut, peer, netif, peer_netif): + local = _host_addresses(dut, netif) + remote = _host_addresses(peer, peer_netif) + routes = [] + try: + for address_family, family in (('inet', 4), ('inet6', 6)): + if address_family not in local or address_family not in remote: + continue + route = _prepare_route( + dut, family, local[address_family], remote[address_family], + netif) + if route: + routes.append(route) + route = _prepare_route( + peer, family, remote[address_family], local[address_family], + peer_netif) + if route: + routes.append(route) + except Exception: + _remove_routes(routes) + raise + if not any(family in local and family in remote + for family in ('inet', 'inet6')): + raise RuntimeError('DUT and peer have no common global address family') + return routes + + +def _remove_peer_key(peer, marker): + # Overwrite authorized_keys only with a complete copy, a failure, e.g. on + # a full disk, must not leave the peer without keys. grep exits with 1 + # when it removed every line, 2 on errors. + command = ( + 'auth="$HOME/.ssh/authorized_keys"; ' + 'test -f "$auth" || exit 0; ' + 'tmp=$(mktemp "${auth}.nipa.XXXXXX") || exit 1; ' + f'grep -vF {shlex.quote(marker)} "$auth" > "$tmp"; ' + 'if [ $? -le 1 ]; then cat "$tmp" > "$auth"; else false; fi; ' + 'status=$?; rm -f "$tmp"; exit $status' + ) + _ssh(peer, command) + + +def _install_peer_key(dut, peer, remote_root): + """Authorize an ephemeral key for SSH from the DUT to the peer.""" + marker = f'nipa-local-{uuid.uuid4().hex}' + remote_key = f'{remote_root}/peer-key' + with tempfile.TemporaryDirectory(prefix='nipa-peer-key-') as temporary: + key = Path(temporary) / 'id_ed25519' + _run(['ssh-keygen', '-q', '-t', 'ed25519', '-N', '', + '-C', marker, '-f', str(key)]) + public_key = key.with_suffix('.pub').read_text( + encoding='utf-8').strip() + command = ( + 'umask 077; mkdir -p "$HOME/.ssh"; ' + 'auth="$HOME/.ssh/authorized_keys"; touch "$auth"; ' + f'grep -Fqx {shlex.quote(public_key)} "$auth" || ' + f'printf "%s\\n" {shlex.quote(public_key)} >> "$auth"' + ) + _ssh(peer, command) + try: + _rsync(key, f'{dut}:{remote_key}') + _ssh(dut, f'chmod 600 {shlex.quote(remote_key)}') + except Exception: + _remove_peer_key(peer, marker) + raise + return marker, remote_key + + +def _validate_assignment(value): + key, separator, setting = value.partition('=') + if not separator or not key or '\n' in value: + raise argparse.ArgumentTypeError('expected KEY=VALUE') + return key, setting + + +def _tree_branch(tree): + completed = _run( + ['git', '-C', str(tree), 'branch', '--show-current'], + capture_output=True, text=True) + branch = completed.stdout.strip() + if not branch: + raise RuntimeError(f'Linux tree is not on a branch: {tree}') + return branch + + +def _resolve_tests(tree, tests): + """Resolve kselftest selectors in a Linux source tree.""" + tree = Path(tree).expanduser().resolve() + root = (tree / 'tools/testing/selftests').resolve() + if not (root / 'run_kselftest.sh').is_file(): + raise RuntimeError(f'not a Linux source tree: {tree}') + + selected = [] + for test in tests: + target, separator, program = test.partition(':') + if not target or (separator and not program): + raise RuntimeError( + f'invalid test {test!r}; expected target[:program]') + target_path = (root / target).resolve() + if root == target_path or root not in target_path.parents: + raise RuntimeError(f'test escapes the selftests tree: {test}') + if not target_path.is_dir(): + raise RuntimeError(f'target does not exist: {target_path}') + if separator: + test_path = (target_path / program).resolve() + if target_path not in test_path.parents: + raise RuntimeError(f'test escapes its target: {test}') + if not test_path.is_file(): + raise RuntimeError(f'test does not exist: {test_path}') + selected.append(test) + + return root, list(dict.fromkeys(selected)) + + +def _select_installed_tests(installed, selectors): + """Expand collection selectors using the installed kselftest list.""" + list_path = installed / 'kselftest-list.txt' + available = [line.strip() for line in list_path.read_text( + encoding='utf-8').splitlines() if line.strip()] + selected = [] + for selector in selectors: + target, separator, _program = selector.partition(':') + matches = [selector] if separator and selector in available else [] + if not separator: + matches = [test for test in available + if test.startswith(f'{target}:')] + if not matches: + raise RuntimeError(f'test was not installed: {selector}') + selected.extend(matches) + return list(dict.fromkeys(selected)) + + +def _stage_selftests(source, tests, destination): + """Install already-built selftests into a temporary runnable tree.""" + targets = list(dict.fromkeys( + test.partition(':')[0] for test in tests)) + make = ['make', '-C', str(source), f'TARGETS={" ".join(targets)}'] + + completed = _run([*make, '-q'], check=False) + if completed.returncode == 1: + raise RuntimeError( + f'selected selftests are not built; run {shlex.join(make)} first') + if completed.returncode: + raise subprocess.CalledProcessError(completed.returncode, make) + + _run([*make, f'INSTALL_PATH={destination}', 'install']) + + +def _worker_problems(stdout): + """Return the sentinels of a worker run a crash or a timeout cut short.""" + return [line for line in stdout.splitlines() + if line.startswith('NIPA DETECTED ')] + + +def _write_worker_env(path, args): + values = { + 'NETIF': args.netif, + 'REMOTE_TYPE': 'ssh', + 'REMOTE_ARGS': args.peer, + 'REMOTE_IFNAME': args.peer_netif, + 'NIPA_SKIP_NET_SETUP': 'yes', + 'NIPA_TEST_TIMEOUT': str(args.timeout), + } + values.update(dict(args.env)) + with open(path, 'w', encoding='utf-8') as fp: + for key, value in values.items(): + fp.write(f'{key}={value}\n') + + +def _local_run(args): + tests_dir, selectors = _resolve_tests(args.tree, args.test) + branch = _tree_branch(args.tree) + + start = datetime.datetime.now(datetime.timezone.utc) + run_name = start.strftime('%Y%m%d-%H%M%S') + output = Path(args.output or f'nipa-results/{run_name}').expanduser().resolve() + output.mkdir(parents=True, exist_ok=False) + remote_root = f'/tmp/nipa-local-{run_name}-{os.getpid()}' + remote_tests = f'{remote_root}/tests' + remote_nipa = f'{remote_root}/nipa' + remote_results = f'{remote_root}/results' + + selected_tests = [] + staging_dir = Path(tempfile.mkdtemp(prefix='nipa-selftests-')) + installed_tests = staging_dir / 'selftests' + worker_status = 0 + worker_stdout = '' + worker_stderr = '' + peer_key_marker = None + remote_peer_key = None + routes = [] + try: + _stage_selftests(tests_dir, selectors, installed_tests) + selected_tests = _select_installed_tests(installed_tests, selectors) + test_list = ''.join(f'{test}\n' for test in selected_tests) + _ssh(args.dut, f'mkdir -p {shlex.quote(remote_tests)} ' + f'{shlex.quote(remote_nipa + "/contest")} ' + f'{shlex.quote(remote_results)}') + routes = _prepare_routes( + args.dut, args.peer, args.netif, args.peer_netif) + peer_key_marker, remote_peer_key = _install_peer_key( + args.dut, args.peer, remote_root) + _rsync(str(installed_tests) + '/', f'{args.dut}:{remote_tests}/') + _ssh(args.dut, + f'rm -f -- {shlex.quote(remote_tests + "/.attempted")} ' + f'{shlex.quote(remote_tests + "/.attempted.tmp")}; ' + f'rm -rf -- {shlex.quote(remote_results)}; ' + f'mkdir -p {shlex.quote(remote_results)}') + _rsync(str(HW_DIR) + '/', f'{args.dut}:{remote_nipa}/contest/hw/') + _rsync(str(ROOT_DIR / 'contest/remote') + '/', + f'{args.dut}:{remote_nipa}/contest/remote/') + _rsync(str(ROOT_DIR / 'contest/__init__.py'), + f'{args.dut}:{remote_nipa}/contest/__init__.py') + + with tempfile.TemporaryDirectory(prefix='nipa-local-') as temporary: + temporary = Path(temporary) + env_path = temporary / 'nic-test.env' + _write_worker_env(env_path, args) + _rsync(env_path, f'{args.dut}:{remote_tests}/nic-test.env') + list_path = temporary / 'kselftest-list.txt' + list_path.write_text(test_list, encoding='utf-8') + _rsync(list_path, + f'{args.dut}:{remote_tests}/kselftest-list.txt') + + command = ( + 'export PATH=/usr/bin:$PATH; ' + 'eval "$(ssh-agent -s)" >/dev/null || exit; ' + "trap 'ssh-agent -k >/dev/null 2>&1' EXIT; " + f'ssh-add {shlex.quote(remote_peer_key)} >/dev/null 2>&1 || exit; ' + f'cd {shlex.quote(remote_nipa)}/contest/hw && ' + f'/usr/bin/python3 hw_worker.py --run-only ' + f'{shlex.quote(remote_tests)} ' + f'--results-dir {shlex.quote(remote_results)}' + ) + completed = _ssh(args.dut, command, capture_output=True, text=True, + check=False) + worker_status = completed.returncode + worker_stdout = completed.stdout + worker_stderr = completed.stderr + print(worker_stdout, end='') + if worker_stderr: + print(worker_stderr, end='', file=sys.stderr) + finally: + (output / 'worker.stdout').write_text(worker_stdout, encoding='utf-8') + (output / 'worker.stderr').write_text(worker_stderr, encoding='utf-8') + try: + (output / 'test-outputs').mkdir(exist_ok=True) + _rsync(f'{args.dut}:{remote_results}/', + str(output / 'test-outputs') + '/') + _run(['scp', *SSH_OPTIONS, + f'{args.dut}:{remote_tests}/.attempted', + str(output / 'attempted.json')]) + except subprocess.CalledProcessError: + print('Warning: some result artifacts could not be fetched', + file=sys.stderr) + _remove_routes(routes) + if remote_peer_key: + try: + _ssh(args.dut, f'rm -f -- {shlex.quote(remote_peer_key)}') + except subprocess.CalledProcessError: + print('Warning: failed to remove the temporary peer key', + file=sys.stderr) + if peer_key_marker: + try: + _remove_peer_key(args.peer, peer_key_marker) + except subprocess.CalledProcessError: + print('Warning: failed to revoke the temporary peer key', + file=sys.stderr) + try: + _ssh(args.dut, f'rm -rf -- {shlex.quote(remote_root)}') + except subprocess.CalledProcessError: + print(f'Warning: failed to remove {remote_root} on {args.dut}', + file=sys.stderr) + shutil.rmtree(staging_dir, ignore_errors=True) + + if worker_status: + raise RuntimeError(f'worker exited with status {worker_status}; ' + f'see {output}/worker.stderr') + + data = { + 'branch': branch, + 'branch_date': start.strftime('%Y-%m-%d--%H-%M'), + 'remote': 'local', + 'executor': args.executor or args.dut, + 'start': start.isoformat(), + 'dut': args.dut, + 'peer': args.peer, + 'test_tree': str(tests_dir), + 'tests': selected_tests, + 'results': parse_results(str(output), '.'), + } + result_path = output / 'result.json' + result_path.write_text(json.dumps(data, indent=2) + '\n', encoding='utf-8') + print(f'Results: {result_path}') + # Like the hardware orchestrator, look for the sentinels of the worker + problems = _worker_problems(worker_stdout) + for problem in problems: + print(f'nipa local: {problem}, the run is incomplete', + file=sys.stderr) + return 1 if problems else 0 + + +def _parser(): + parser = argparse.ArgumentParser(prog='nipa local') + subcommands = parser.add_subparsers(dest='command', required=True) + + run = subcommands.add_parser('run', help='run prepared kselftests on a DUT') + run.add_argument('--tree', required=True, help='prepared Linux source tree') + run.add_argument('--dut', required=True, help='DUT SSH destination') + run.add_argument('--peer', required=True, + help='peer SSH destination, reachable from this host ' + 'and the DUT') + run.add_argument('--netif', required=True, help='DUT test interface') + run.add_argument('--peer-netif', required=True, help='peer test interface') + run.add_argument('--test', action='append', required=True, + metavar='TARGET[:PROGRAM]', + help='kselftest collection or program; repeatable') + run.add_argument('--env', action='append', type=_validate_assignment, + default=[], metavar='KEY=VALUE', + help='extra worker or kselftest environment variable') + run.add_argument('--timeout', type=int, default=1200) + run.add_argument('--output') + run.add_argument('--executor') + run.set_defaults(func=_local_run) + return parser + + +def main(argv=None): + args = _parser().parse_args(argv) + try: + return args.func(args) + except (OSError, RuntimeError, subprocess.CalledProcessError, + json.JSONDecodeError) as error: + print(f'nipa local: {error}', file=sys.stderr) + return 1 diff --git a/contest/hw/tests/test_local.py b/contest/hw/tests/test_local.py new file mode 100644 index 00000000..b61ed18f --- /dev/null +++ b/contest/hw/tests/test_local.py @@ -0,0 +1,176 @@ +# SPDX-License-Identifier: GPL-2.0 + +import json +import os +import subprocess +import tempfile +import unittest +from pathlib import Path +from unittest import mock + +from contest.hw.local import _prepare_route, _remove_peer_key, _remove_routes, \ + _resolve_tests, _select_installed_tests, _worker_problems + + +class TestLocalRoutes(unittest.TestCase): + def test_installs_and_removes_host_route(self): + current = mock.Mock(stdout=json.dumps([{ + 'dev': 'eth0', + 'prefsrc': '2001:db8::10', + }])) + desired = mock.Mock(stdout=json.dumps([{ + 'dev': 'eth1', + 'gateway': 'fe80::1', + 'prefsrc': '2001:db8:1::1', + }])) + installed = mock.Mock(returncode=0) + removed = mock.Mock(returncode=0) + + with mock.patch('contest.hw.local._ssh', + side_effect=[current, desired, installed, removed]) as ssh: + route = _prepare_route( + 'dut', 6, '2001:db8:1::1', '2001:db8:2::1', 'eth1') + _remove_routes([route]) + + self.assertEqual( + ssh.call_args_list[2].args, + ('dut', 'ip -6 route add 2001:db8:2::1/128 via fe80::1 ' + 'dev eth1 src 2001:db8:1::1 proto static')) + self.assertEqual( + ssh.call_args_list[3].args, + ('dut', 'ip -6 route del 2001:db8:2::1/128 via fe80::1 ' + 'dev eth1 src 2001:db8:1::1 proto static')) + + def test_preserves_matching_route(self): + current = mock.Mock(stdout=json.dumps([{ + 'dev': 'eth1', + 'prefsrc': '2001:db8:1::1', + }])) + with mock.patch('contest.hw.local._ssh', return_value=current) as ssh: + route = _prepare_route( + 'dut', 6, '2001:db8:1::1', '2001:db8:2::1', 'eth1') + + self.assertIsNone(route) + ssh.assert_called_once() + + +class TestLocalPeerKey(unittest.TestCase): + KEYS = ('ssh-ed25519 AAAA admin\n' + 'ssh-ed25519 BBBB nipa-local-1234\n') + + @staticmethod + def _ssh_here(home): + def ssh(_host, command, **kwargs): + return subprocess.run(['sh', '-c', command], + env={**os.environ, 'HOME': str(home)}, + check=kwargs.get('check', True)) + return ssh + + def _authorized_keys(self, temporary): + path = Path(temporary) / '.ssh' / 'authorized_keys' + path.parent.mkdir() + path.write_text(self.KEYS, encoding='utf-8') + return path + + def test_removes_only_the_temporary_key(self): + with tempfile.TemporaryDirectory() as temporary: + path = self._authorized_keys(temporary) + with mock.patch('contest.hw.local._ssh', + side_effect=self._ssh_here(temporary)): + _remove_peer_key('peer', 'nipa-local-1234') + + self.assertEqual(path.read_text(encoding='utf-8'), + 'ssh-ed25519 AAAA admin\n') + self.assertEqual([entry.name for entry in path.parent.iterdir()], + ['authorized_keys']) + + @unittest.skipIf(os.geteuid() == 0, 'root ignores file permissions') + def test_keeps_keys_when_no_copy_can_be_written(self): + with tempfile.TemporaryDirectory() as temporary: + path = self._authorized_keys(temporary) + # authorized_keys stays writable, its directory does not + path.parent.chmod(0o500) + try: + with mock.patch('contest.hw.local._ssh', + side_effect=self._ssh_here(temporary)), \ + self.assertRaises(subprocess.CalledProcessError): + _remove_peer_key('peer', 'nipa-local-1234') + finally: + path.parent.chmod(0o700) + + self.assertEqual(path.read_text(encoding='utf-8'), self.KEYS) + + +class TestLocalWorkerStatus(unittest.TestCase): + def test_sentinels_report_an_incomplete_run(self): + stdout = ('[1/2] Running drivers/net/perf:devmem.py\n' + 'NIPA DETECTED SYSTEM CRASH\n' + 'NIPA DETECTED TEST TIMEOUT\n') + + self.assertEqual(_worker_problems(stdout), [ + 'NIPA DETECTED SYSTEM CRASH', 'NIPA DETECTED TEST TIMEOUT', + ]) + self.assertEqual(_worker_problems('[1/2] Running x\n'), []) + + +class TestLocalTestPaths(unittest.TestCase): + def test_resolves_tests_in_linux_tree(self): + with tempfile.TemporaryDirectory() as temporary: + tree = Path(temporary) + root = tree / 'tools/testing/selftests' + (root / 'run_kselftest.sh').parent.mkdir(parents=True) + (root / 'run_kselftest.sh').write_text('', encoding='utf-8') + target = root / 'drivers/net/perf' + target.mkdir(parents=True) + (target / 'iperf.py').write_text('', encoding='utf-8') + (target / 'devmem.py').write_text('', encoding='utf-8') + selected = [ + 'drivers/net/perf:iperf.py', + 'drivers/net/perf:devmem.py', + ] + + test_root, tests = _resolve_tests(tree, selected) + + self.assertEqual(test_root, root) + self.assertEqual(tests, selected) + + def test_resolves_and_expands_collection(self): + with tempfile.TemporaryDirectory() as temporary: + tree = Path(temporary) + root = tree / 'tools/testing/selftests' + (root / 'run_kselftest.sh').parent.mkdir(parents=True) + (root / 'run_kselftest.sh').write_text('', encoding='utf-8') + target = root / 'drivers/net/perf' + target.mkdir(parents=True) + + test_root, selectors = _resolve_tests( + tree, ['drivers/net/perf']) + + installed = tree / 'installed' + installed.mkdir() + (installed / 'kselftest-list.txt').write_text( + 'drivers/net/perf:devmem.py\n' + 'drivers/net/perf:iperf.py\n' + 'drivers/net/hw:ethtool.sh\n', encoding='utf-8') + tests = _select_installed_tests(installed, selectors) + + self.assertEqual(test_root, root) + self.assertEqual(tests, [ + 'drivers/net/perf:devmem.py', + 'drivers/net/perf:iperf.py', + ]) + + def test_rejects_invalid_test_identifier(self): + with tempfile.TemporaryDirectory() as temporary: + tree = Path(temporary) + root = tree / 'tools/testing/selftests' + root.mkdir(parents=True) + (root / 'run_kselftest.sh').write_text('', encoding='utf-8') + + with self.assertRaisesRegex(RuntimeError, + r'target\[:program\]'): + _resolve_tests(tree, ['drivers/net/perf:']) + + +if __name__ == '__main__': + unittest.main() diff --git a/nipa b/nipa new file mode 100755 index 00000000..2aa580fd --- /dev/null +++ b/nipa @@ -0,0 +1,19 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 + +"""Small command-line entry point for local NIPA tools.""" + +import sys + +from contest.hw.local import main as local_main + + +def main(): + if len(sys.argv) >= 2 and sys.argv[1] == 'local': + return local_main(sys.argv[2:]) + print('usage: nipa local run ...', file=sys.stderr) + return 2 + + +if __name__ == '__main__': + sys.exit(main()) From 1e98482630ac5bbab5ec570ae4de2c699f539d5b Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Tue, 29 Sep 2026 08:49:06 -0700 Subject: [PATCH 18/18] contest: hw: add local result viewer Build metric API responses from local result.json files. Serve them through the existing performance UI without a database. Signed-off-by: Stanislav Fomichev --- README.rst | 13 +++ contest/hw/local.py | 173 +++++++++++++++++++++++++++++++++ contest/hw/tests/test_local.py | 45 ++++++++- nipa | 2 +- ui/metric-table.js | 3 +- 5 files changed, 232 insertions(+), 4 deletions(-) diff --git a/README.rst b/README.rst index 6e8c6f1d..151985c9 100644 --- a/README.rst +++ b/README.rst @@ -104,6 +104,19 @@ Results are written to ``$PWD/nipa-results/``. Use ``target:program`` to select one program or ``--output`` to change the results directory. +How to view +~~~~~~~~~~~ + +Serve all results under ``$PWD/nipa-results``:: + + cd $nipa + ./nipa local view + +Pass a path after ``view`` to select another directory or one result. The +server listens on ``[::]:44109`` and prints the URL without opening a browser. +History is grouped by Linux branch and executor. Regression evaluation uses +up to five prior successful values. + Structure ========= diff --git a/contest/hw/local.py b/contest/hw/local.py index ce5ed410..9b7bb9dd 100644 --- a/contest/hw/local.py +++ b/contest/hw/local.py @@ -4,21 +4,28 @@ import argparse import datetime +import http.server import json import os from pathlib import Path import shlex import shutil +import socket +import socketserver import subprocess import sys import tempfile +import urllib.parse import uuid +from contest.backend.result_metrics import evaluate_regressions, \ + extract_result_metrics from contest.hw.lib.result_parser import parse_results HW_DIR = Path(__file__).resolve().parent ROOT_DIR = HW_DIR.parent.parent +UI_DIR = ROOT_DIR / 'ui' SSH_OPTIONS = [ @@ -28,6 +35,18 @@ ] +class _ThreadingTCPServer(socketserver.ThreadingTCPServer): + allow_reuse_address = True + + +class _ThreadingTCPServerV6(_ThreadingTCPServer): + address_family = socket.AF_INET6 + + def server_bind(self): + self.socket.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_V6ONLY, 1) + super().server_bind() + + def _run(command, **kwargs): print('+', shlex.join(str(part) for part in command)) kwargs.setdefault('check', True) @@ -394,6 +413,154 @@ def _local_run(args): return 1 if problems else 0 +def _find_results(path): + path = Path(path).expanduser().resolve() + if path.is_file(): + return [path] + return sorted(path.rglob('result.json')) + + +def _metric_rows(result_paths): + overview = [] + details = [] + artifact_roots = {} + next_id = 1 + + loaded = [] + for path in result_paths: + with open(path, encoding='utf-8') as fp: + loaded.append((path, json.load(fp))) + loaded.sort(key=lambda item: (item[1]['start'], str(item[0]))) + + for artifact_id, (path, data) in enumerate(loaded): + artifact_roots[str(artifact_id)] = path.parent + for row in extract_result_metrics(data): + link = row.get('link') + if link and not urllib.parse.urlparse(link).scheme: + relative = link.lstrip('./') + row['link'] = f'/artifacts/{artifact_id}/{relative}' + branch = data['branch'] + if data['remote'] == 'local': + branch = f"{branch}-{data['branch_date']}" + row.update({ + 'id': next_id, + 'branch': branch, + 'branch_date': data['branch_date'], + 'remote': data['remote'], + 'executor': data['executor'], + 'start': data['start'], + }) + details.append(row) + overview.append({key: value for key, value in row.items() + if key != 'full_json'}) + next_id += 1 + + overview = evaluate_regressions(overview) + evaluated = {row['id']: row['short_json'] for row in overview} + for row in details: + row['short_json'] = evaluated[row['id']] + return overview, details, artifact_roots + + +def _handler(results_path): + class LocalHandler(http.server.SimpleHTTPRequestHandler): + def __init__(self, *args, **kwargs): + super().__init__(*args, directory=str(UI_DIR), **kwargs) + + def _send_json(self, data): + payload = json.dumps(data, separators=(',', ':')).encode() + self.send_response(200) + self.send_header('Content-Type', 'application/json') + self.send_header('Content-Length', str(len(payload))) + self.end_headers() + self.wfile.write(payload) + + @staticmethod + def _load_metrics(): + return _metric_rows(_find_results(results_path)) + + def do_GET(self): # pylint: disable=invalid-name + parsed = urllib.parse.urlparse(self.path) + path = parsed.path + params = urllib.parse.parse_qsl( + parsed.query, keep_blank_values=True) + if path in ('/', '/metric-table.html') and \ + not any(key == 'data' for key, _value in params): + params.extend([ + ('data', '/local/result-metrics.json'), + ('detail-data', '/local/result-metric-details.json'), + ]) + location = '/metric-table.html?' + urllib.parse.urlencode(params) + self.send_response(302) + self.send_header('Location', location) + self.send_header('Content-Length', '0') + self.end_headers() + return + if path == '/local/result-metrics.json': + overview, _details, _artifact_roots = self._load_metrics() + self._send_json(overview) + return + if path == '/local/result-metric-details.json': + _overview, details, _artifact_roots = self._load_metrics() + self._send_json(details) + return + if path.startswith('/artifacts/'): + _overview, _details, artifact_roots = self._load_metrics() + parts = path.split('/', 3) + root = artifact_roots.get(parts[2]) if len(parts) > 3 else None + if root is None: + self.send_error(404) + return + requested = (root / urllib.parse.unquote(parts[3])).resolve() + if root != requested and root not in requested.parents: + self.send_error(403) + return + if requested.is_dir(): + requested = requested / 'stdout' + if not requested.is_file(): + self.send_error(404) + return + self.send_response(200) + self.send_header('Content-Type', self.guess_type(str(requested))) + self.send_header('Content-Length', str(requested.stat().st_size)) + self.end_headers() + with open(requested, 'rb') as fp: + shutil.copyfileobj(fp, self.wfile) + return + super().do_GET() + + return LocalHandler + + +def _local_view(args): + result_paths = _find_results(args.results) + if not result_paths: + raise RuntimeError(f'no result.json files found under {args.results}') + overview, _details, _artifact_roots = _metric_rows(result_paths) + if not overview: + raise RuntimeError('no testcase metrics found in the selected results') + + handler = _handler(args.results) + server_class = _ThreadingTCPServerV6 if ':' in args.host else \ + _ThreadingTCPServer + with server_class((args.host, args.port), handler) as server: + host, port = server.server_address[:2] + url_host = f'[{host}]' if ':' in host else host + query = urllib.parse.urlencode({ + 'data': '/local/result-metrics.json', + 'detail-data': '/local/result-metric-details.json', + }) + url = f'http://{url_host}:{port}/metric-table.html?{query}' + print(f'Serving {len(overview)} metric attempts from ' + f'{len(result_paths)} runs') + print(url) + try: + server.serve_forever() + except KeyboardInterrupt: + print() + return 0 + + def _parser(): parser = argparse.ArgumentParser(prog='nipa local') subcommands = parser.add_subparsers(dest='command', required=True) @@ -416,6 +583,12 @@ def _parser(): run.add_argument('--output') run.add_argument('--executor') run.set_defaults(func=_local_run) + + view = subcommands.add_parser('view', help='view local result.json files') + view.add_argument('results', nargs='?', default='nipa-results') + view.add_argument('--host', default='::') + view.add_argument('--port', type=int, default=44109) + view.set_defaults(func=_local_view) return parser diff --git a/contest/hw/tests/test_local.py b/contest/hw/tests/test_local.py index b61ed18f..be912376 100644 --- a/contest/hw/tests/test_local.py +++ b/contest/hw/tests/test_local.py @@ -8,8 +8,8 @@ from pathlib import Path from unittest import mock -from contest.hw.local import _prepare_route, _remove_peer_key, _remove_routes, \ - _resolve_tests, _select_installed_tests, _worker_problems +from contest.hw.local import _metric_rows, _prepare_route, _remove_peer_key, \ + _remove_routes, _resolve_tests, _select_installed_tests, _worker_problems class TestLocalRoutes(unittest.TestCase): @@ -172,5 +172,46 @@ def test_rejects_invalid_test_identifier(self): _resolve_tests(tree, ['drivers/net/perf:']) +class TestLocalMetricStore(unittest.TestCase): + def test_reads_result_json_without_database(self): + with tempfile.TemporaryDirectory() as temporary: + path = Path(temporary) / 'result.json' + path.write_text(json.dumps({ + 'branch': 'local-test', + 'branch_date': '2026-09-16--12-00', + 'remote': 'local', + 'executor': 'dut', + 'start': '2026-09-16T12:00:00+00:00', + 'results': [{ + 'group': 'selftests-net', + 'test': 'example-py', + 'result': 'pass', + 'link': '.', + 'results': [{ + 'test': 'case-one', + 'result': 'pass', + 'metrics': [{ + 'name': 'throughput', + 'type': 'scalar', + 'value': 10.0, + 'host': 'local', + 'summarize': 'p50', + }], + }], + }], + }), encoding='utf-8') + + overview, details, roots = _metric_rows([path]) + self.assertEqual(len(overview), 1) + self.assertEqual(overview[0]['executor'], 'dut') + self.assertEqual( + overview[0]['branch'], + 'local-test-2026-09-16--12-00') + self.assertEqual( + overview[0]['short_json'][0]['summary']['p50'], 10.0) + self.assertEqual(details[0]['full_json'][0]['name'], 'throughput') + self.assertEqual(roots['0'], path.parent) + + if __name__ == '__main__': unittest.main() diff --git a/nipa b/nipa index 2aa580fd..d76f4481 100755 --- a/nipa +++ b/nipa @@ -11,7 +11,7 @@ from contest.hw.local import main as local_main def main(): if len(sys.argv) >= 2 and sys.argv[1] == 'local': return local_main(sys.argv[2:]) - print('usage: nipa local run ...', file=sys.stderr) + print('usage: nipa local {run,view} ...', file=sys.stderr) return 2 diff --git a/ui/metric-table.js b/ui/metric-table.js index 7473e14a..06fad1ea 100644 --- a/ui/metric-table.js +++ b/ui/metric-table.js @@ -13,7 +13,8 @@ function metric_case_name(run) function metric_run_date(run) { - return run.branch_date.slice(0, 10); + return run.remote === 'local' ? run.branch_date : + run.branch_date.slice(0, 10); } function metric_detail_data_source()