Vault using helm chart crash loop

We have been experience a recurrent crash over the last few weeks. It’s seems that some error occurs while vault is running and terminate the pod, there is not any log that could guide us the only thing that we notice is an up rising in memory and CPU consumption of the pods. At some point the crash stops and the pod continues running.

You can notice the picks in the charts and the abrup falls are when the pods finally crash.

You should review the pod logs. Note that kubectl logs has a --previous option.