Alerts generated when 'Unmanage' period ends

The point of unmanage is that it completely stops monitoring. Therefore, when unmanage ends the first thing SolarWinds does is collect a full set of statistics. If one of those statistics breaches a threshold, you are going to get an alert, the same as if it had been monitored all along and JUST breached the threshold.

Even using the "Mute" work around (), you would get the alert, because once you un-mute the relevant query properties have changed.

The only way to block alerts would be further upstream - meaning that Solarwinds sends it's alert to some other system (your ticketing server, etc) and THAT system sends out alerts. Then you could implement a blackout feature at that level. So that during the blackout, Solarwinds is sending alerts but the ticket system sees the blackout and blocks it. And once the blackout is over, Solarwinds isn't going to send ANOTHER alert.

However, in every place I've been what I just described is a BUG, not a FEATURE. The immediate request from all teams is "I need you to RE-check the device when it comes out of blackout so I know if it's still down".

Which alert is fired?

After unmanage period alerting engine fire alert because that node fulfill trigger condition.

You can try prolong unmanage period or you can set "Do not trigger until this condition exists for more than" to 2 min for example in affected alert.