chore(pipeline-perf): add container CPU-seconds metric

[cijothomas]

Summary

Adds a cgroup-based CPU-seconds metric to the pipeline-perf report:
cpu_seconds — the CPU time a component consumes during the observation window,
derived from a new cumulative container.cpu.time gauge (docker stats
total_usage).

Unlike the existing container.cpu.usage (an instantaneous, wall-clock-normalized
utilization ratio), cpu_seconds is an absolute work measure integrated over
the observation window. In a rate-limited test, fixed input work => fixed
CPU-seconds, which makes it a clean, stable regression signal. It needs no
perf/PMU, so it is available on GitHub-hosted runners as well as dedicated
hardware.

Changes

• Docker monitor (docker_component.py): record a cumulative
  container.cpu.time gauge (seconds) alongside the existing
  container.cpu.usage.
• Report (integration_report_logs.yaml): add a component_cpu_time table
  (MAX(value) - MIN(value) over the observation window) and emit a cpu_seconds
  row in gh_actions_benchmark.

Validation

Three label-triggered runs on the dedicated bare-metal runner (full suite).
cpu_seconds run-to-run CoV (df-engine):

Scenario	CoV
OTLP-ATTR-OTLP	0.05%
OTLP-ATTR-OTAP	0.09%
OTAP-ATTR-OTAP	0.19%
OTAP-ATTR-OTLP	0.04%
OTLP-BATCH-OTLP	0.11%
OTAP-BATCH-OTAP	0.06%

All under 0.2% — a solid, low-variance signal for regression detection. Also
verified via a DuckDB dry-run of the report SQL and python3 tools/sanitycheck.py.

Notes / scope

• This is intentionally minimal: one absolute CPU-work metric. It complements the
  existing container.cpu.usage/throughput metrics without adding overlapping
  per-log or hardware-counter columns.
• On GitHub-hosted runners cpu_seconds carries ~1% run-to-run noise (shared-VM
  frequency scaling / neighbors), so it is suitable there only as a coarse gate
  for large regressions; precise tracking belongs on the dedicated/nightly runs.

Kept as draft.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 86.51%. Comparing base (601c953) to head (17009b2).
⚠️ Report is 4 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3515      +/-   ##
==========================================
+ Coverage   86.25%   86.51%   +0.25%     
==========================================
  Files         767      782      +15     
  Lines      302274   313105   +10831     
==========================================
+ Hits       260729   270876   +10147     
- Misses      41021    41705     +684     
  Partials      524      524              

Components	Coverage Δ
otap-dataflow	87.53% <ø> (+0.26%)	⬆️
query_engine	89.57% <ø> (ø)
otel-arrow-go	52.45% <ø> (ø)
quiver	92.19% <ø> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[opentelemetry-pr-dashboard]

Pull request dashboard status

Status last refreshed: 2026-07-22 19:57:43 UTC.

• Waiting on: Author
• Next step: Move out of draft to request review.

This automated status or its linked feedback items may be incorrect. If something looks wrong, please report it with the result you expected.