So, we finished upgrading the organisation to SP3 this morning. We had performed the upgrade to SP3 for the Client Access, Hub Transport and Edge Transport servers last week (as per MS best practice) which was a fairly simple process – you can just run the upgrade through a nice GUI by running the setup.exe. This week’s upgrade of all the clusters was still simple but does require a bit of different process. Hell, they’re clusters – what do you expect..?
This process in this post is taken from http://technet.microsoft.com/en-us/library/bb676320(EXCHG.80).aspx (I’m assuming you’ve already done all CAS, Hub and Edge servers)…
For the purpose of this process, I’ll use the names Node1 and Node2 as the active and passive node changes throughout the exercise.
Upgrade the passive node (Node2) to SP3 by running the following command:
Restart the passive node (Node2).
Once it’s back online, log onto Node2 (still passive) and run the following command to stop the cluster. Mailboxes will go offline at this point:
Stop-ClusteredMailboxServer CLUSTERNAME -StopReason "Upgrade to SP3"
Now move the cluster to Node2 making it the active node using the following command:
Move-ClusteredMailboxServer CLUSTERNAME -TargetMachine NODE2 -MoveComment "Upgrade to SP3"
Now the cluster is running on Node2 we need to upgrade the CMS. This is done with the following command. When this is complete it will automatically bring the cluster back online on Node2:
Now we need to upgrade Node1 (now passive) to SP3 using the command we used in step 1:
Restart the server once complete.
Your CCR cluster is now upgraded to SP3. Pat yourself on the back.
I found that the outage (step 3 to step 5) for the CMS upgrade was about 5-10 minutes in my environment. This will obviously differ depending on the speed of your machines and the location of the source files.
Hope this helps someone out there…