This is a new one for me. I have lots of hours on Hyper-V 2008 R2 and 2012 R2. I have lots of hours on fail over clusters with HPE Proliant DL380 GenX servers with MSA storage. I am aware of the Broadcom driver bug that requires disabling VMQ. I built a new Windows 2016 Hyper-V fail over cluster on HPE Proliant DL380 Gen9 servers and MSA storage. Everything working perfectly, except networking.
The VMs on the cluster CAN obtain DHCP addresses form a server elsewhere on the network (or I can set them statically - it doesn't matter how the VMs are addressed) and they also CAN access the corporate network AND the internet as expected. But nothing can access the VMs, cannot access shares, cannot RDP, cannot ping. Also, none of the VMs can access other VMs on the same cluster and exhibit the same behavior.
Just so you can grasp the strangeness of this, imagine virtual servers VM1 and VM2 on the cluster, and NONVM1 outside the cluster. VM1 can access NONVM1, shares, RDP, ping, etc, as expected. But, even though NONVM1 knows VM1 exists due to DNS resolution only, it cannot access it at all, no shares, no RDP, no ping. Also, VM1 CANNOT access VM2 on the same cluster and vice versa.
All drivers and firmware are up to date. I have tried and assumed it was VMQ settings, but this does not appear to be the cause. The Cluster Validation runs and passes everything.
I'm at a total loss to explain this one.
Any ideas, and thank you in advance.