I have 4 identical nodes. By identical I mean even the timing on the RAM are the same across all four nodes. These are Proliant DL360 G6 servers which I have previously used in a clusters. OS is Server 2008 R2 SP1 Core, drivers and firmware
are all identical. Each server has an additional fiber HBA card and quad port NIC. Drivers and firmware is identical on each card. System firmware is the same. I built the server using HP smartstart 8.7 and did not install HP's network
config utility since it causes issues.
All nodes are attached to an HP EVA8400, each node can see the two LUNs I am beginning with.
When I used the failover cluster manager to validate the configuration I get the following on one node:
An error occurred while executing the test.
There was an error verifying the IP configuration.
Unable to connect to Server03 via WMI. This may be due to networking issues or firewall configuration on Server03.
I can run a WMI test from another node and get a response from Server03 indicating the WMI is working properly. For example the command to get maxclockspeed. I have disabled the firewall on all nodes. All networking settings are identical. Have a management NIC, heartbeat, and livemigration NIC. The remaining three ports are for VMs. These are all one the same switch stack (have the heartbeat and live migration isolated). Every address is pingable from each node. There are no hardware firewalls.
So I try to add the machines to VMM 2012. I add them and attempt to create a cluster and get the following error:
An error occurred while executing the test.
Invalid class
This happens on the Gathering inventory data for servers running Hyper-V portion. And it bombs out. I tried again removed the troubled Server03 and just using Server01 and Server02. Same error.
When using the failover manager all other tests pass.
I built a hyper-v cluster when R2 first came out and I swear I did not have this much trouble. I can't use 2012 because our EVA won't support it so I have to figure this out.