✨ Refactor telemetry and metrics handling to enhance observability. Remove TLS configuration from telemetry initialization, replacing it with an insecure option for HTTP. Update OpenTelemetry endpoint handling to normalize URLs. Introduce new metrics for publish failures categorized by message type. Enhance Grafana dashboards with detailed descriptions and additional metrics for better monitoring insights. Update documentation to reflect changes in metrics and observability features.

2025-11-19 14:53:58 +08:00
parent c687fdcde8
commit f01dd3bf1f
13 changed files with 224 additions and 238 deletions
@@ -183,8 +183,9 @@ func (p *promRecorder) RecordProcessingResult(ctx context.Context, status, categ
 }

 func (p *promRecorder) RecordPublishFailure(ctx context.Context, category string) {
-	// Prometheus metrics currently only expose failures via caatsm_failures_total,
-	// so we record the publisher failure there.
+	// Record publish failure with category label for detailed tracking.
+	obsmetrics.RecordPublishFailure(category)
+	// Also record to legacy failure counter for backward compatibility.
 	obsmetrics.RecordFailure("publisher")
 }