Running NPM 12.3, HA mode with 1 pool (just a primary and secondary server) on Windows 2012 R2
I've noticed during HA fail overs from the primary to the secondary that the Universal Device Pollers we have set up do not poll/run on the secondary (meaning: when we are running on the secondary, all of UDP charts/graphs in the web ui are blank, when we fail back to the primary and they reappear and there is a hole in the data during the fail over time). These fail overs are initiated by NPM (because of issues on the primary) and not manually forced fail overs.
Does anyone know if these specific pollers *should* be running on the secondary when failed over to the secondary or if this is expected behavior/known issue? Just wanted to bounce this off of the community first if anyone has seen this issue before opening a ticket as I can't find any info saying they shouldn't work, but also nothing explicitly stating they should 
.
