When Polling Engine server restarts, nodes go to unknown state and Alert actions executed

Question

I have an alert for for when a node goes to a "Down" state for hundreds of nodes. I have three polling engines (PE.) When the services restart, or the PE is patched and restarted alert notifications are sent out for the nodes reporting to or being polled by the PE. The Reset actions are triggered a few minutes after that.

Has anyone else experienced this and have a solution? I can think of some options but they all seem painful to implement so your suggestions are wanted. Thanks!

sturdyerde · Answer

I'm running the last version prior to 2019.4, maybe minus a hotfix or two. Will check the MSQM queues. Thanks for the tip and the success link!

m_roberts · Answer

This is something that used to happen often on older versions, but I have not seen much of this in recent versions.

Things to check:

1. Do you have any queues in your MSMQ list on the APEs?

2. When you restart the APE server/services do you notice any nodes change their status? if they do, this indicates your Pub/Subs could do with resetting (Success Center )

sturdyerde · Answer

Great question. I have also noticed that whenever a polling engine restarts, all agent-managed nodes are alerted as down or unknown, followed by a reset notification. It's....messy.