Would it be possible to mark only CPU-usage related "trouble" outages as maintenance across all of our Linux machines?
We have decided to remove this monitor as the "Load" monitor gives us a better idea of actual CPU stress (based on CPU wait time and actual load on CPU) but it outages from it still remain... Can we mark these as maintenance using the Site24x7 API? We have a lot of Linux machines being monitored and we cannot go through one-by-one and mark as maintenance.
Thanks!
Dear Alex,
Site24x7's IT Automation can help you in marking a server as maintenance when the threshold is breached.
In the threshold profile that you have configured for you Linux servers, each attribute has an automation step. This means that whenever a threshold is breached for a server say CPU crosses 70% continuously for 2 times, you can configure to run an automation for it. In this case you need to select the "mark as maintenance" automation.
You can also write a custom script and execute it as an automation in case you want do more than just marking it as maintenance.
Here are the steps to achieve this
1. In server threshold profile > Under the attribute CPU Select the automation you want to trigger. I have created an mark as maintenance for 10 min automation.
2. To create an mark-as-maintenance (MaM) automation goto Admin > IT Automation > click on the + > select types as MaM. Once this is done you can set this in the above step.
Once this is set, whenever a server reached 70% CPU and stays at 70+ % for 2 polls, the "10 min maintenance" will be executed for that server.
Does this solve your case? If you any clarifications, please get back to us.
-Jasper
Product Manager, Site24x7