Move PerformanceMonitor sqlite3 I/O off the event loop

[groupthinking]

Run synchronous sqlite3 connect/INSERT/commit/close in a thread so PerformanceMonitor metric/alert writes don’t stall request coroutines.

[linear-code]

GRV-228

[github-actions]

👋 Thanks for opening this issue! It has been automatically labeled for triage.

[coderabbitai]

⚠️ Possible Duplicate Issue(s)

• perf: PerformanceMonitor runs blocking sqlite3 I/O on the event loop #1195

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!

[groupthinking]

Already resolved on main — closing as complete

This was filed on 2026-08-01, the same day the fix merged. The tracking issue was never
linked to the PR that implemented it, so it stayed open. The work described here is
already shipped; leaving it open risks somebody re-implementing it.

Fix: 26b2a7cf8 (#1196) — perf(performance-monitor): move blocking sqlite3 I/O off the event loop
Code: src/youtube_extension/backend/services/performance_monitor.py

All blocking sqlite3 access in PerformanceMonitor is routed through asyncio.to_thread, keeping the event loop responsive during metric writes.

Verification performed

1. Commit is on main — git merge-base --is-ancestor 26b2a7cf8 origin/main exits 0.

2. Mechanism present in current origin/main — confirmed by reading
   git show origin/main:src/youtube_extension/backend/services/performance_monitor.py, not just the merge record.

3. Regression suite shipped with the fix passes:

   pytest tests/unit/test_performance_monitor.py
   # 119 passed
   

No further work required.

_{Verified by an automated triage pass. The whole async-fix cluster
(#1191, #1192, #1198, #1199, #1201, #1206) was re-run together: 442 passed, 0 failed.}