fix: resolve traceroute race condition for fast-responding nodes (#2364)

[Yeraze]

Summary

Fixes #2364 — traceroutes showed "No response received" for nearby/fast-responding nodes despite the response being visible in logs and packet monitor.

Root Cause

recordTracerouteRequest() for PostgreSQL/MySQL fired an un-awaited async IIFE to insert the pending traceroute record. For nearby nodes, the traceroute response arrived before the DB insert completed:

1. sendTraceroute() → sends packet, calls await recordTracerouteRequestAsync()
2. But the async wrapper called the inner method without await → returned immediately
3. Fast response arrives → processTracerouteMessage() runs
4. DB insert for the pending record is still in flight
5. Frontend reads stale DB state → shows "No response received"

Fix

Made recordTracerouteRequest() properly async — removed the fire-and-forget IIFE and directly awaits the DB operations. The async wrapper now correctly awaits the inner method.

Files Changed

File	Change
src/services/database.ts	Make recordTracerouteRequest async, remove un-awaited IIFE, await in async wrapper

Test plan

• 3052 tests pass, 0 failures
• SQLite path unchanged (already synchronous)
• PostgreSQL/MySQL path now properly awaited

🤖 Generated with Claude Code

[github-actions]

✅ Great fix! Making this method properly async resolves the race condition. The previous fire-and-forget IIFE meant fast traceroute responses could arrive before the pending record was inserted into the database.

Consider adding a brief comment here explaining why the await is critical:

Suggested change

	async recordTracerouteRequest(fromNodeNum: number, toNodeNum: number): Promise<void> {
	async recordTracerouteRequest(fromNodeNum: number, toNodeNum: number): Promise<void> {
	// NOTE: This MUST be awaited to prevent race conditions with fast-responding nodes
	const now = Date.now();

[github-actions]

✅ Perfect removal of the problematic pattern. The fire-and-forget IIFE was the root cause of the race condition. Now the database operations are properly awaited before the method returns, ensuring the pending traceroute record exists when fast responses arrive.

[github-actions]

✅ Critical fix! This change from fire-and-forget to properly awaited call is essential. The async wrapper pattern throughout the codebase expects the inner method to be awaited, and this was the missing piece causing the race condition.

This now properly aligns with the multi-database architecture requirement that "ALL database methods in DatabaseService must be async".