Apache Kafka is a distributed event-streaming platform that publishes, stores, and delivers records through topics. In an event flow, topic inspection helps establish whether a message propagated between services. section 13.1.1
Use Kafka inspection to investigate time-based anomalies:
- Check whether the expected record reached the relevant topic. section 13.1.1
- Compare topic evidence with distributed traces and application logs; one tool rarely gives the full picture. section 13.1.1
- Treat missing records as clues, then investigate the last service that completed a downstream action. section 13.1.1
The missing one-euro transaction
A payment begins at a POS device, passes through an API gateway and payment service, and should be audited and executed elsewhere. Checking POS logs followed by the Kafka topic can reveal whether the auditing service failed to receive or continue the event. section 13.1.1
Figure 13.2 The engineer checked the POS logs followed by the Kafka topic. The clues they found revealed the possibility that the problem is in the auditing service side.
Inspect the POS logs and Kafka topic together to localize where the payment event stopped propagating. section 13.1.1
