Traefik 3.7.13 Health Check

Traefik Traefik v3 (latest)
1

Hello, for some test yesterday, I have added two dead services in a route with a health check, like this into the dynamic config.

    healthCheck:
      interval: 10s
      timeout: 3s

I can quickly see the both are dead and no more load balancing (due to the health check), it's working.

The test is good and I leave. I'm looking today and the three are down.

No idea why but a restart keep the three down and a compose down/up solve the issue.

How could I avoid this to keep the service on, same with only one server.

Thanks for your help (hope this message is at the right place)

2

Adding the "unhealthyInterval" field and setting it to one hour resolves the issue. If I use the same value as for the interval 30s, the problem comeback quickly.

What is the optimal value for you?