We operate a single-DAG environment where each mailbox database has two passive copies in addition to an active (mounted) copy. In our environment we're seeing periodic "spikes" in copy queue length for mailbox databases that clear themselves out
relatively quickly. I'd like to correlate those "spikes" to some event; our working theory is they come from a sudden uptick in user activity. Maybe a user is performing large-scale operations in their mailbox of some kind, for instance.
Trouble is, I'm not sure how to monitor for that user activity on a per-database level. Of course I know I can use ExMon to look at user connections to databases, but how do I correlate that to actual activity on a mailbox database in a monitoring sense?
As we don't have a way to predict when these "spikes" in queue length occur I would need to monitor and track this activity over an extended period of time and then correlate it with other monitoring we have in place (Solar Winds) for copy queue
lengths.
Thanks in advance for your help.
