add CPU throttling metrics from cgroup cpu.stat

[JayiceZ]

fix VictoriaMetrics/VictoriaMetrics#10430

---

Summary by cubic

Expose CPU throttling metrics from cgroup cpu.stat for the current process. Works with cgroup v1 and v2 and only emits metrics when the cgroup CPU controller is enabled (addresses #10430).

• New Features
  • Added metrics: process_cgroup_cpu_periods_total, process_cgroup_cpu_throttled_periods_total, process_cgroup_cpu_throttled_seconds_total.
  • Auto-detects v1/v2 cpu.stat and parses nr_periods, nr_throttled, and throttled time; converts to seconds and skips metrics if the CPU controller isn’t enabled.

^{Written for commit 085d533. Summary will update on new commits.}

[cubic-dev-ai]

2 issues found across 1 file

Prompt for AI agents (all issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="process_metrics_linux.go">

<violation number="1" location="process_metrics_linux.go:474">
P2: writeProcessCpuThrottleMetrics dereferences ctms without guarding against the nil return from getCgroupCpuStats, which can panic when no cgroup CPU stats are available.</violation>

<violation number="2" location="process_metrics_linux.go:476">
P2: `process_cgroup_cpu_throttled_seconds_total` reports non-second units (micro/milliseconds). Convert throttled_time to seconds (or store microseconds consistently and divide by 1e6 when writing) to avoid unit mismatch.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[cubic-dev-ai]

P2: process_cgroup_cpu_throttled_seconds_total reports non-second units (micro/milliseconds). Convert throttled_time to seconds (or store microseconds consistently and divide by 1e6 when writing) to avoid unit mismatch.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At process_metrics_linux.go, line 476:

<comment>`process_cgroup_cpu_throttled_seconds_total` reports non-second units (micro/milliseconds). Convert throttled_time to seconds (or store microseconds consistently and divide by 1e6 when writing) to avoid unit mismatch.</comment>

<file context>
@@ -403,6 +420,62 @@ func readPSITotals(cgroupPath, statsName string) (uint64, uint64, error) {
+	}
+	WriteGaugeUint64(w, "process_cgroup_cpu_periods_total", ctms.nrPeriods)
+	WriteGaugeUint64(w, "process_cgroup_cpu_throttled_periods_total", ctms.nrThrottled)
+	WriteGaugeUint64(w, "process_cgroup_cpu_throttled_seconds_total", ctms.throttledTime)
+}
+
</file context>

[codecov]

Codecov Report

❌ Patch coverage is 49.18033% with 31 lines in your changes missing coverage. Please review.
✅ Project coverage is 87.02%. Comparing base (58ce999) to head (085d533).
⚠️ Report is 4 commits behind head on master.

Files with missing lines	Patch %	Lines
process_metrics_linux.go	49.18%	26 Missing and 5 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #112      +/-   ##
==========================================
- Coverage   88.68%   87.02%   -1.67%     
==========================================
  Files          12       12              
  Lines        1388     1449      +61     
==========================================
+ Hits         1231     1261      +30     
- Misses        106      132      +26     
- Partials       51       56       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[cubic-dev-ai]

No issues found across 1 file