perf(server): reduce Windows WMI polling

[nateEc]

What Changed

• Replace the per-process performance counter lookup with two bulk WMI queries and an in-memory PID map.
• Slow background resource sampling from every 5 seconds to every 15 seconds.
• Add focused coverage that prevents the N+1 WMI query pattern from returning.

Why

The diagnostics poller issued one additional WMI query for every process on every sample. On Windows this caused sustained WmiPrvSE CPU usage and repeated failed IDProcess queries.

Closes #4024

Checklist

• Focused diagnostics tests pass
• Server typecheck passes
• Targeted lint and formatting pass

---

Note

Low Risk
Diagnostics-only changes on Windows WMI and slower background sampling; no auth, data, or API contract changes beyond approximate CPU rollup timing.

Overview
Cuts Windows process diagnostics cost by replacing per-process WMI performance lookups with two bulk Get-CimInstance queries and a PID→CPU map in the PowerShell one-liner, instead of an N+1 -Filter query per process.

Background resource sampling interval moves from 5s to 15s; rollup tests now expect higher cpuSecondsApprox because that metric uses the new interval.

A regression test asserts a single powershell.exe spawn per sample, exactly two Get-CimInstance calls, no -Filter, and use of $perfById.

^{Reviewed by Cursor Bugbot for commit e4fe092. Bugbot is set up for automated code reviews on this repo. Configure here.}

Note

Reduce Windows WMI polling frequency and eliminate per-process CIM queries

• Rewrites the PowerShell command in readWindowsProcessRows (ProcessDiagnostics.ts) to fetch all Win32_PerfFormattedData_PerfProc_Process data in a single Get-CimInstance call, building a $perfById lookup instead of issuing a filtered query per process.
• Increases SAMPLE_INTERVAL_MS in ProcessResourceMonitor.ts from 5,000 ms to 15,000 ms, tripling the time between samples.
• Behavioral Change: CPU metrics on Windows now reflect a single bulk snapshot rather than per-process filtered queries; processes missing from the perf snapshot fall back to 0% CPU.

^{Macroscope summarized e4fe092.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 218558bf-cdf3-4b67-9e32-679ec6b034a5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Needs human review

The polling interval change from 5s to 15s is a meaningful runtime behavior modification affecting monitoring responsiveness, not just an internal optimization. This warrants human review to ensure the reduced granularity is acceptable.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3-code]

thanks! closing as superseded by #2679, merged in 49c0d96. that change removed the powershell/wmi diagnostics poller and old sampling path entirely in favor of the native resource monitor.