pkg/containerwatcher/v2/tracer_manager.go (2)

141-161: ⚠️ Potential issue | 🟠 Major

Unconditional 30s wait even when procfs tracer is absent or disabled.

The select block at lines 154–158 waits for ProcfsScanInterval (default 30s) regardless of whether the procfs tracer was found or started. If the procfs tracer doesn't exist in the map or is disabled, this is a pointless 30-second stall on every startup.

Move the wait inside the conditional block so it only applies when the tracer was actually started.

Proposed fix

 func (tm *TracerManager) startProcfsTracer(ctx context.Context) error {
 	if tracer, exists := tm.GetTracer(utils.ProcfsEventType); exists {
 		delete(tm.tracers, utils.ProcfsEventType)
 		if tracer.IsEnabled(tm.cfg) {
 			logger.L().Info("Starting procfs tracer before other tracers")
 			if err := tracer.Start(ctx); err != nil {
 				return fmt.Errorf("starting procfs tracer: %w", err)
 			}
+			// Wait for procfs scan to complete before starting eBPF tracers.
+			// This delay ensures container/process data is ready for enrichment.
+			select {
+			case <-time.After(tm.cfg.ProcfsScanInterval):
+			case <-ctx.Done():
+				return ctx.Err()
+			}
 		}
 	}
 
-	// Wait for procfs scan to complete before starting eBPF tracers.
-	// This delay ensures container/process data is ready for enrichment.
-	select {
-	case <-time.After(tm.cfg.ProcfsScanInterval):
-	case <-ctx.Done():
-		return ctx.Err()
-	}
-
 	return nil
 }

---

15-29: ⚠️ Potential issue | 🟡 Minor

thirdPartyTracers has no registration method and remains unused.

The thirdPartyTracers slice is initialized in the constructor but never populated. All tracers, including third-party tracers, are registered via RegisterTracer() which adds them to the tracers map. This makes the thirdPartyTracers field dead code, as stopThirdPartyTracers() iterates over an always-empty slice. Either remove this unused field or populate it during tracer registration if separate handling of third-party tracers is intended.