Community
- Command Central
- MVP Program
- Monthly Mission
- Blogs
- Groups
- Events
- Media Vault
Products
- Observability
- Network Management
- Application Management
- IT Security
- IT Service Management
- System Management
- Database Management
Content Exchange
- SolarWinds Platform
- Server & Application Monitor
- Database Performance Analyzer
- Server Configuration Monitor
- Network Performance Monitor
- Network Configuration Manager
- SQL Sentry
- Web Help Desk
Free Tools & Trials

Excluding servers on a scheduled reboot from alerting?

dkandola

I have a generic alert set up for our "server" estate. We have over 500 servers within our server estate.

An email alert is sent when a server does down. (device_type = server and node_status = down).

Every night, we have approximately 10 servers that re-boot.

What I would like to do is avoid these from alerting when they go down during the re-boot phase.

I am aware that this can be achieved by 'umanaging' the device.

However, is there an alternative way within the alerts to avoid the servers that re-boot from alerting?

Find more posts tagged with

Accepted answers

All comments

gregatkins

I'm off work for a couple of days so I'm trying to see visualize the entire alert application in my head .

You should be able to specify to alert on a node being down or a node being rebooted. You could also add a condition to meet that if the node is rebooted then don't send an alert out. Of course, if you have some very tight polling values it could be a race of how fast the server reboots and is able give data to Solarwinds vs. how quickly your polls are going out. I've had that happen before on some older servers where we had to customize the polling values to give it some time to come back on..even when the alert was set not to go off on reboots. One of the Solarwind techs I spoke to said to try and avoid using the alert supression tab and use the main alert tab by supressing alerts with added conditions.

You may have to change it over to an advance alert . The only time you should unmanage a device, in my opinion, is when it is taken out of production or will be offline for an extended period.

Hope this helps you some. Play around with the advanced alerts if you get a chance, it gives you a ton of options and you won't break anything

aLTeReGo

The simplest solution would be to require the node to be down for a sustained period of time before being alerted. That time would be greater than is reasonable for a server reboot. A more complex method would be to write a script using the Orion SDK to execute as part of the server shutdown script that unmanages the node in Orion. Then do the reverse, executing a startup script to re-manage the node when it comes back up.

nrms

Also, you can change the time of day that an alert is active, so if you have a set time they reboot at, change your alert to only be active outside this time.

Of course, this works best if your reboots are just before or just after midnight. If it happens at, for example, 9pm you would need 2 alerts - 1 active from midnight to 8:55pm and anotherfrom 9:15pm to 11:59pm (or other suitable times). If the reboot happens at 11:30pm, just make your existing alert active from midnight to 11:25pm.

rjnicholson

Just use the Unscheduling Management Utility on the Orion Server. You pick the nodes and applications you want and set a period of time you want to have them go into an unmanaged state. Once you build this and save the profile you use Windows task manager on the Orion server to run the task you just created. It will run a script and unmanage the node in the DB/Orion for the selected period of time. I do this for all of our Critical Business applications and nodes during our Saturday maintenance window of 6pm - 12pm. I just find out what time the servers reboot from the server team and build the Utility and Task manager around that time frame. I usually give the unmanaged time about 5 minutes extra time frame to let the node come full back online from the reboot.

Let me know if you need any help with this. Just make sure you log into the Orion server with an Admin account to build the Unmanaged schedule and the Windows Task in Task manager. The task manager in Windows requires you to run the task from an Admin account, and this account also needs to have DB access to NetPerfMon.