We recently upgraded our Hyper-V Cluster to 2016 and are in the process of deploying new SQL Clusters within it using the new Shared VHDS in 2016. We are looking for some best practices on how to configure the job to do Application Aware Processing on the SQL Instances.
The SQL Cluster will be Server 2016 running SQL 2016 in a 2 node Active/Passive configuration. In the past on Hyper-V 2012 with a Guest Server 2012 SQL Cluster using the Shared VHDX's what we were doing was to run the backups of all the volumes from the Passive node and only backup the C: drive of the Active node, this just reduced the stress on the active node during the backup. But 2012 didn't allow us to do Application Aware Processing so all we were getting was crash consistent backups of the volumes. We aren't sure if that same method will work with Application Aware Processing now that we can do that with 2016.
Has anyone done something similar yet or know of any KB articles on best practices for this type of arrangement? We are getting close to finishing up our Test Cluster, so we may need to be the guinea pigs here and just test out a few different methods to see what works best.