We're seeing some really strange behaviour with a 2 node cluster on azure 2019 datacenter. We didn't see the issue right away but at some point it started happening and now we can repeat it.
We have an azure shared disk that we assign as a cluster shared volume in failover cluster manager. If we reboot one of the nodes upon starting up again windows explorer becomes unresponsive for quite some time. Interestingly powershell also is unresponsive (can't even type a command into it) until windows explorer becomes responsive. We've launced powershell using task manager. However launching a command window from task manager does not have the delay.
We've removed all roles from the cluster. Removed the software that was installed and formatted the CSV drive so it was all clean.
If we remove the disk as a CSV and leave it in available disks and reboot we do not get the delay. If we add it back on as a CSV we get the delay again. We can repeat this as much as we want.
If we bounce both nodes at once it takes up to 45 minutes for explorer and powershell to become active again. Doing the same with no CSV we have no issue.
I can't see anything in the event logs that is indicating the problem. It's a real strange phenomena.
I'd say it was a one off but we had this issue previously and decided to just redeploy from scratch. Everything worked fine for a day or 2 and then it started again.
We're pretty much at the end of things we can try and I'm wondering if anything has seen anything similar or if there is anything else we can look at.
