Hi All,

I've got a 3 node Hyper V cluster running Server 2012 R2 and live migration is failing between 2 of the nodes. Live migrations work back and forth from node 2 to all other nodes fine. Live migrations from node 3 work to all other nodes, Node 1 can live migrate to node 2 but fails when migrating to node 3 and gets the following error:

Log Name: System

Event ID: 21502

Live migration of 'Virtual Machine XXXX' failed.

Virtual machine migration operation for 'XXXX' failed at migration source 'Node 1'.

The Virtual Machine Management Service failed to establish a connection for a Virtual Machine migration with host 'Node 3': A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond. (0x8007274C).

The Virtual Machine Management Service failed to authenticate the connection for a Virtual Machine migration at the source host: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond. (0x8007274C).

Failed to send data for a Virtual Machine migration: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond. (0x8007274C).

Looking around at other support articles on this it looks like this error code (0x8007274C) points to Virtual Machine Manager Service not listening on port 6600 but have confirmed it is listening on all nodes running netstat -ano | findstr 6600 and live migration does work on all nodes (just not one way between these two). Quick migration works fine from Node 1 to Node 3.

All Nodes are running the exact same hardware with the following nic configuration:

Broadcom BCM57800 NetXtreme 11 10 GigE cards.
Dual 10gbe ports teamed (traffic untagged at switch)
2 vEthernet adapters configured on teamed nic, one for management and one solely for migrations, each on different subnets.

Obviously this a tricky one to troubleshoot but if someone can give me any tips or tools I can use to further troubleshoot this it would be appreciated.

Cheers

Exact same issue here... been trying everything I could find without success except a reboot.

Have you had a chance to reboot your nodes to see if it helps?

I did reboot the affected hosts and found the problem changed, now node 1 was able to live migrate to node 3 but not the other way. Still no issues with node 2 which can live migrate between all the other hosts. Our storage is connected via Fibre so ethernet is just for LAN connectivity and migrations.

Did you make sure that all 3 nodes have permission via role delegation for Live Migrations and File System access?  Follow the steps found here for more details(note the steps turn on access to everything, I would choose only the 2 roles I mentioned plus any others you deem needed.)

http://www.virtuallycloud9.com/index.php/2013/11/server-2012-r2-lab-guide-part-4-step-by-step-hyper-v-over-smb/

Thanks Tommy but this didn't make a difference, I tried adding delegation for the cifs and Microsoft Virtual System Migration Service for the other nodes, when this didn't work I added all the services but still no difference. I didn't think this config was necessary when you were using a Hyper V cluster? Original settings on the Node AD accounts was "Do not trust this computer for delegation" and Live migration does work between some of the nodes.

I changed the live migration settings today to use the Management network and Live migration worked between these nodes so looks network related. I'll go through the network settings today and see if I can find any discrepancies. I'll keep digging to try and find out what the 0x8007274C error code specifically points to for a failure.