We have an on-premise service Fabric 5 node Cluster spread on 5 separate VMs. There are 5 Upgrade Domains. The current version of Service Fabric is 6.4.637.9590. We are trying to upgrade to 6.5.676.9590 and facing issues in doing so.
When the upgrade is started using the Start-ServiceFabricClusterUpgrade command through the powershell CLI, the entire cluster goes down. Initially I can run the Get-ServiceFabricClusterUpgrade command immediately after the upgrade starts and it comes back with a "In Progress" status. However thereafter the cluster goes down and I am not even able to connect to the cluster to get the upgrade status.
We have tried this twice and both times this has failed and we have had to restore the VMs from previous snapshots.
We have tried upgrading the cluster once with the application running and also without any applications deployed in the cluster and both times we have seen the same behaviour.
Can you please help as we do not want to recreate the cluster from scratch everytime there is a new version released.