Automated way to suppress alerts during Polling Engine weekly reboot

Hello,

In our environment we have weekly reboots for all four of our polling engines. Unfortunately when these servers are doing their weekly reboots the nodes being polled by them appear down so alerts are sent out. Is there a way to suppress these alerts? So for the only way is to schedule an alert disable during the reboots but I would prefer if there was a way to stop alerts going out based on the polling engine that is being rebooted.

For example: Say Polling Engine1 (PE100) is being rebooted. I would like all nodes polled by PE100 to go into unreachable status or unmanaged. Thanks!

Find more posts tagged with

polling engine

Alerts

Accepted answers

All comments

SW_United

This has a hot topic for many other Thwack posts, I've found this post probably the closest, with thanks to @bobmarley

[mention:4fcf6b1a560c4f60b49dc6a9af7ecc10:fb5d84b10a5745448a7a45dafc1faa43]

Hope this helps

stevenstadel

Scheduled Maintenance is a feature of 2025.2 RC. Release Notes If you want a workaround until you get that in place you can use the Manage Nodes view to sort by polling engine and then schedule the muting of the nodes.

Cheesecrisp89

Thanks for bringing this to my attention. I guess Im more looking for the reboot/down status of my polling engine to be the trigger for these nodes to be unmanaged then a time based schedule. Thanks for the help though.

bourlis

Wait., I think I'm missing something. Are you saying when you reboot your SW server(s) that an alert is sent for a managed node that wasn't rebooted? Or are you saying that an alert is being sent for the SolarWinds server itself? These are two completely different issues.

If I'm reading your issue correctly, you're saying the alerts are from managed nodes that are being managed that's kicked off during the SW server(s) reboots. If I'm correct the problem is more than likely caused by the way the alarm is built or how the nodes are managed (i.e, ICMP versus agent and/or server-initiated or agent-initiated communication).

If this is the case, it would be fairly easy to isolate the root cause. Starting with one agent change its communication method. I'm guessing your agents are server-initiated which would most likely cause SW to think their down if it can't reach them. If I'm correct change it to agent-initiated, then sit back and wait for the next round of reboots. If you don't get an alert from that server then that's your issue.

If you still get the alert next look at the alert itself. Your alerts should be very explicit like 'Scope of alert = Node type = Windows' and then 'Trigger condition is Node status = down'. If you have anything other than trigger condition down, then consider changing it. Lastly, double check the condition must exist settings. If your environment allows it, set it to 20 minutes. 20 minutes should cover most Windows reboots. But if you have Cisco nodes then it should be set to 5 minutes.