We are having upcoming activity for Hardware Firmware upgrade and have a DAG setup also for exchange 2016. I am looking to understand/clarify if we can do vmotion of exchange node, OR this will cause a DB failover which is not needed.
because for two nodes, I need to do this 4 times:
- shift node 1 from physical host (perform firmware update)
-return node 1 to same physical host
- shift node 2 from physical host (perform firmware update
--return node 2 to same physical host
kindly share your experiences for this case for vmotion of exchange node part of DAG
I have setup several DAGs for different customers and I usually set the “samesubnetdely“ to 2 seconds. The default is 1 second which can cause unwanted failovers.
You can change this by using this command:
cluster /prop samesubnetdelay=2000
You can check the success by using
This will result in a list of all parameters.
Look at the following link on page 36, it's about the configuration of vSphere vMotion for Exchange DAG
Thank you for response. Some places I have seen other then parameter you have specified, people are suggesting for these also, what is your opinion:
cluster /cluster:yourDAGname /prop SameSubnetThreshold=10
cluster /cluster:yourDAGname /prop CrossSubnetDelay=4000
cluster /cluster:yourDAGname /prop CrossSubnetThreshold=10