"Verify webhook with unavailable endpoints" false alarm?

[yuvals]

Hi, I'm getting an insight on my cluster:
"Verify webhook with unavailable endpoints", which points me out to this doc link, yet I see my service is up and running, for the last 24 hours, why I'm getting this insight?

[debolek]

The "Verify webhook with unavailable endpoints" insight indicates that there is a webhook in your GKE cluster that is pointing to an endpoint that is not reachable.. Ensure that the webhook configuration is correct, the service and endpoints are healthy, and there are no network policies or firewall rules blocking traffic to the webhook service.

[yuvals]

Hi @debolek, thanks for the response.

I completely understand the insight content, yet the service which handles the webhook is completely healthy, and there is no sign of any unhealthy or pending pods.

Given those circumstances, why I'm getting this insight?

[jerbia7]

Running into the same behaviour, any solution to get rid of those annoying notifications ?

[Adam_H]

I have rolled out a new GKE Autopilot cluster today and am also getting this message. The pods are running and the service is also available.

It's confusing

[jholguin]

Same here, I am getting this error and is confusing, all pods / services are reachable.

[jayeshmahajan]

• Startup Race Conditions: At cluster boot or during upgrades, the webhook might not be ready briefly, which gets cached by GKE's insights engine.

• Delayed Propagation: Even after the service is healthy, the insight doesn’t always auto-clear.

• Autopilot Constraints: GKE Autopilot may restrict or delay certain startup behaviors (like DNS resolving early), which briefly makes webhooks appear unreachable.

• Ephemeral Failures: Temporary DNS lookup or service endpoint sync issues can cause this warning even if the cluster auto-recovers quickly.