Ingress health check not following ReadinessProbe

[matti]

with:

        readinessProbe:
          httpGet:
            path: /health
            port: 8080

the google health check is still /

Also confirmed by @Gogoro:

" When I create an ingress pointing to a service, which again points to a pod it just keeps hitting / instead of the path I defined in readinessProbe and livenessProbe. I can see in the logs that the pod itself checks itself easily, but the healthcheck goes ham on /."

In this discussion issue: #42

[nicksardo]

PTAL at this comment: #241 (comment)

Also, this is only retrieved on new GCP healthcheck objects. If you're adding the readiness probe post ingress setup, nothing will change.

[JorritSalverda]

I have the same issue when changing the readinessProbe on existing applications I'm switching to GCE ingress. If the rolling update is slow an old readinessProbe is sometimes used. Would be nice to follow changes in the probe in order to update the health check. However since a service can resolve to pods with different readiness probes it might be useful to override the health check path in the ingress object with an annotation, although I'm not sure how that would work if ingress routes to multiple service. An annotation on the service instead?

[nicksardo]

@JorritSalverda , Unfortunately, due to legacy behavior, that cannot be done. See #39

[irvifa]

Have the same problem here, actually, I need to create my own health check path in the GCP UI after updating my ingress. I already used k8s version 10.2.

So what can we do @nicksardo ?

[irvifa]

Seems like it's related to #42