3 ms·

You may have misunderstood me. The case I'm talking is, the node reports Ready, but the pod itself is not functioning properly. One common issue we have is the

by webo 10y ago

You may have misunderstood me. The case I'm talking is, the node reports Ready, but the pod itself is not functioning properly. One common issue we have is the pod gets stuck in a restart loop (for whatever reason, including starvation of resources). k8s just keeps restarting it for days on that node, instead of simply rescheduling it after X restarts or some other condition. https://github.com/kubernetes/kubernetes/issues/13385 https://github.com/kubernetes/kubernetes/issues/13385

shaklee3 10y ago
Isn't that what custom liveliness probes are for?
- webo 10y ago
  No- liveness probe just restarts the pod if the check fails, doesn't kill it.
  - amouat 10y ago
    I think a failed liveness will restart the container, not the whole pod.
    - webo 10y ago
      You're right. We always have one container per pod so I didn't fully think about this.
      - shaklee3 10y ago
        Isn't that because the pod's restart policy is set to restart? What if you set that to off? Does it fail the whole pod?