Health check from other service configured with retry policy to config server can hang.

We encountered some strange behavior when we have a service, which is configured to contact the config-service with a retry policy (for instance try for 30 minutes) with regards to that service health check page.

Looks like the same retry policy is used for updating the health information of the config service in that service. However, when the config service is down, this check is retrying and retrying. At that time, the /health page of the service hangs and doesn't show.

We use external automated monitoring on /health to identify the health status. In the above case, it marks the service as down, while it isn't.

I'd guess the health check towards the config server should not take into account the configured retry policy and just show DOWN instantly when it is not there.

Kind regards,
Wouter

Start by reproducing the interaction between the service /health endpoint, the config-server health check, and a long configured retry policy while the config server is unavailable. Trace the health-check and retry-policy handling in the Spring Cloud Config client; it is done when /health responds promptly and reports the config service as down instead of hanging.