What happened: Application down due to runaway Redis cache worker consuming all CPU.
Root cause: The cache optimization service worker maintains Redis memory efficiency by periodically analyzing and compacting cached query results. During routine optimization, the worker entered an infinite loop, consuming 100% CPU cycles and starving all application processes. The worker automatically restarted by design (supervised process), immediately re-triggering the same infinite loop and continuing the outage.
Fix: Disabled automatic restart on the service worker supervisor, killed the runaway process, verified app stability, then manually restarted the worker with additional safeguards.
Prevention: Implement circuit breaker with auto-shutdown on high resource usage.