Adding DAG node gets erver-side DAG administrative operation failed, with Error 0x71a (The remote procedure call was cancelled) from cli_RpccAddNodeToCluster
I am creating an Exch 2010 DAG in my test lab.
I can create the DAG and the first node without issue however, adding the second DAG node gets:
A server-side DAG administrative operation has failed. The operation failed with message: Error 0x71a (The remote
procedure call was cancelled) from cli_RpccAddNodeToCluster
Error 0x71a (The remote procedure call was cancelled) from cli_RpccAddNodeToCluster
upon executing:
Add-DatabaseAvailabilityGroupServer -Identity 'TEST11-DAG1' -MailboxServer 'TEST-11-VM4'
Both DAG nodes (original and one trying to add) are hyper-v machines.
I believe the cluster fail over feature is installed on both.
Both hyper-v nodes have a public DHCP Public Virtual Network cards and a Private heartbeat.
The initial DAG virtual node has a static ip as well so three virtual cards.
I am running Windows 2008 r2 on both.
I disabled the firewalls on both nodes:
Features installed on both:
Cluster Failover service
Windows process activation
Branch Cache
.net 3.5.1
remote server admin tools -active directory
powershell ISE
Roles
Webserver IIS
fILE Services
Any ideas? Or suggestions on where to look?
All Replies
- Hello,
Are both nodes running Windows 2008 Enterprise?
Did you try configuring the DAG from the second node? I initially had this problem with my VMs when I tried to configure the DAG from my Primary Exchange 2010 Server.
When I re-ran DAG configuration from the server that failed to add, and I added it to the DAG, the process completed successfully.
I'm not sure why this was necessary or what was different with running from the Secondary server. Both servers had their firewalls turned off.
I hope this helps,
Trevor - Adding a machine to the DAG involves several steps, and some of those steps have changed from the Beta to RTM. RTM is more robust, and tries to print better error messages.
There is an RPC (Remote Procedure Call) that's made to a machine currently in the DAG, and that's the one that performs the operation.
The 71a error means that the RPC has timed out -- it did not return after a pre-defined amount of time. 5 minutes. This is done to prevent hangs on one machine from tying up the other indefinitely. Unfortunately it turns out that in the real world this operation sometimes really does take longer than 5 minutes (as an example, sometimes DNS updates are very slow).
As a work-around, add the machine 'test-11-vm4' manually to the cluster using the cluster admin tools. Then re-run add-dagserver. If the Join operation still fails, then at least you have a GUI that should have a more meaningful error than 71a (timeout). Don't forget to re-run add-dagserver, because joining a node to the cluster and not running add-dagserver can result in some strange behaviours.
-martin
Legal Stuff: This posting is provided "AS IS" with no warranties, and confers no rights. - Did you run the verification tests when you added the new server from ClusterManager? If Yes, what was the outcome?
/m.
