In analyzing a performance issue I can see that on a 2 node Availability Group Failover Cluster with synchronous commit to a read-intent only secondary show different numbers of schedulers inside SQL looking at sys.dm_os_sys_info.
As an example we have one set where one node shows 51 schedulers and the other node shows 28. This isn't a matter of the Primary and the Secondary because I did a failover on a pair that showed 11 and 12 schedulers. Those numbers did not change after the failover.
Can anybody explain what the differences and how to correct these?
Thanks
Chris