Clustering can be confusing and for some it may look like smoke and mirrors with how it all work’s and hangs together. I recently was putting together a demo environment with a 2 node Windows Server Failover Cluster (WSFC). During the build of the cluster it seemed to all go well except on looking more closely, Node 2 in the cluster was being reported as offline by the Failover Cluster Manager. Node 2 was up and available and I could ping the server but it was having issues communicating with Node 1 through the Cluster.
Due to not having a lot of time to get my demo environment up and running I decided to destroy the cluster and start a fresh build. The destroy cluster command did not work completely and only removed Node 1 from the cluster, leaving the cluster in a precarious state. The cluster was not online, because Node 2 could not be contacted, Node 1 was no longer a member of the cluster.
I found myself in a position of working out how I can cleanup this mess without having to rebuild both nodes in my cluster to get my demo environment back on track. The Cluster Service on Node 2 was offline and I was not able to start the service, which with it being offline I was unable to destroy the cluster. I tried the following PowerShell command to force a cleanup of the cluster “Get-Cluster –Name [ClusterName] | Remove-Cluster –Force –CleanupAD“. This made no difference for my situation.
After a little more persistence I was able to cleanup the mess on Node 2 with using the “Clear-ClusterNode [ClusterNodeName] –Force” PowerShell command. Once I had run this command I was able to rebuild the cluster using both Nodes 1 & 2 again.
Remember, the use of these types of commands should not be used in production environments unless absolutely necessary and you are fully aware of what is happening in your environment. My Issue was because I was working in Azure in an IaaS environment and clustering is slightly different to on-premise Clusters.