In the process of adding another node to the Amazee CH1 OpenShift cluster an outdated X.509 certificate was used by mistake (a node with the same name was part of the cluster for a few weeks in 2018). Shortly after adding the node to the cluster, monitoring reported the wrong certificate. During the recovery effort one other compute node was impacted and all pods scheduled on that node were forcefully terminated, leading to brief application outage of websites running on that node. Redeploying the certificates in question on all nodes resolved the issue.

To avoid such issues from recurring in the future we will ensure not to reuse hostnames.