I have a 2 node 2012 failover cluster with appox. 20 VMs shared among them. We are seeing massive file access slow downs on our virtual file servers when users are copying big files to the file server. I was wondering if there was a tech note that could has disk I/O best practices in a cluster environment. Would I start looking at counters at host level or VM level? I have tried to even the I/O load by spreading the VHDX files over multiple luns. CPU/Memory/Network Utilization looks fine. Seems to be a disk I/O issue, but how do I verify?
Thanks for your time