A workload with a proxy sidecar fails startup probes even though the application container itself is healthy.
Startup probe keeps failing because the endpoint depends on a sidecar that becomes healthy only after the probe budget is exhausted
The main container is fine, yet startup never succeeds because the readiness path depends on another container or sidecar that comes online later.
Scenario
What to check first
- Identify the primary failure signal in the Cross-Container Health Dependency scenario.
- Separate visible symptoms from the underlying technical dependency.
- Describe the safest recovery path and the follow-up prevention work.
Checking checklist
- Summarize the current impact and the last known change.
- Collect direct evidence from logs, runtime state, and configuration before changing anything.
- Separate immediate recovery from permanent prevention work.
Recovery and prevention
Validate whether the probe endpoint depends on another container's readiness before increasing thresholds again.
Questions worth viewing together
Community-field Kubernetes probe problem inspired by Stack Overflow patterns where startup probes depended on a sidecar path that was not ready in time. A startup probe can fail forever if its endpoint relies on a sibling container that is not ready in time.
Teams often blame the main container when the health endpoint actually depends on the sidecar lifecycle.
Multi-container pods make health semantics more complex than single-container probe patterns suggest.
Similar cases seen in the field