03 Feb 2025 11:52 AM
Hello everyone,
I have a cloud-native Golang application deployed in Kubernetes, and I’m using automated distributed tracing (without OpenTelemetry) to monitor the complete request flow. My application publishes messages to GCP Pub/Sub, but I’ve noticed that after a message is published, I lose visibility into the subsequent processing – specifically, I can’t track the Pub/Sub message handlers as part of the end-to-end trace.
I suspect this issue arises because Dynatrace loses the trace/span ID that is automatically injected before the message is published. As a result, when the handler processes the message, Dynatrace cannot associate it with the original trace.
Has anyone encountered this issue before? Is there a recommended approach to propagate the tracing context (i.e., the trace/span ID) so that Dynatrace can track asynchronous operations like GCP Pub/Sub message handling? Any configuration tips or best practices would be greatly appreciated.