Many customers (even after trained) do not put DAG nodes into maintenance mode and simply install updates on a node and reboot causing a database failover to occur. They then do the remaining servers in the cluster usually ending up with all databases residing on a single server.
In Exchange 2016 CU2 there is a new shiny feature which "automatically fails back" the database to the preferred server based on Activation Preference. This is known as "PreferenceMoveFrequency".
After all your Exchange 2016 servers (or more technically the Primary Active Manager) have been upgraded to Exchange 2016 CU2, you will have a new DAG property called PreferenceMoveFrequency.
What this switch does is define a frequency (measured in time) when the Microsoft Exchange Replication service will rebalance the database copies by performing a lossless switchover that activates the copy with an ActivationPreference of 1.
How cool is that... automatic failback to preferred members.
Now when I build a new Exchange cluster for a customer, after I leave and close out the project I can be sure that the databases will automatically fail back to the preferred server after the customer installs Windows Updates on cluster nodes. This is important as it keeps the load across the cluster balanced.
To set this feature, simply use the following PowerShell command:
Set-DatabaseAvailabilityGroup -Identity DAG01 -PreferenceMoveFrequency ([TimeSpan]::Zero)
Need specialised Exchange consulting in Perth? Contact Avantgarde IT Services on 08 9468 7575