Specify replica=2 in the storage class and deploy the pod with kind:deploy. It operates normally.
Afterwards, a problem occurred with the container within the pod. At this time, if we delete the pod, the pod is redistributed to other nodes. PV also has volume locality and is well connected to the target volume.
I have a question.
We monitor the containers within the pod with a container liveness probe.
If there is a problem with the probe, kubctl will continue to restart only the container.
When a container problem occurs, is there a way to terminate not the container but the pod itself and force redeployment to another node?
I have a question.
We monitor the containers within the pod with a container liveness probe.
If there is a problem with the probe, kubctl will continue to restart only the container.
When a container problem occurs, is there a way to terminate not the container but the pod itself and force redeployment to another node?
This is not really related to LINBIT SDS / Piraeus. This is just generally how Kubernetes works. You would need some additional controller that evicts the Pod once it crashes, but I don’t think there is a built-in way to achieve that. You may want to ask this question in the Kubernetes forums.