Back to practice
Production Incident Lab

Kafka Local Cache: Four Failures Before Lunch

You are on call for a Spring Boot service that materializes configuration from Kafka into every pod. Keep the service correct through four escalating incidents.

5 minutes 4 incidents Middle
Incident 1 of 4
Systems stabilized: 0
INC-001 SEV-2

Evidence

    Relevant configuration
    Your call

    Choose the safest production response. Number keys work too.

    Want the engineering behind the answers?

    Read the complete production pattern

    The deep dive covers compacted topics, per-pod consumer groups, fixed startup targets, readiness gating, tombstones, and the operational limits of this design.

    Read the article