After racking my brain on how to alert on an application monitor that was applied to a group of servers, I came up with the following solution. I'm sure there is a better way, but not exactly sure how to do it - so any suggestions on how to make this better would be great.
The goal was (is) to alert the responsible parties when a component within the application monitor went into a state other than "up" or "running." One server had a few extra services that were being monitored that had to be up. These services (for simplicity's sake) were being monitored on the other two, but I had marked that it was ok if these services were found to be not running. So, in the primary section, I put the two servers that had the non-running service, and then in the secondary section, I set the third server that had to have the services running. The screen shots are below...
Alert Primary Section:
Secondary Alert Section:
In the alert section, I set it to alert on particular instances of the alert, and designated the first two servers that did not have the SQL services running on it. I then used the application with status formatted set to is not equal to the up status of all the running components (except for the SQL services, which read not running).
Then came the scary part...setting the advanced alert part (haven't used that much). In the secondary section I used the same logic for the scope, but designated the SQL server as the only instance. I then used the same logic for the trigger condition, but setting it to up/running status of the SQL server
SIDE NOTE: I'm glad I wrote this, as I see now that I have to fix the trigger condition of the secondary section so that it reads is NOT equal to....score 1 for Thwack QA....
Application Monitor:
For reference purposes, here is the processes and services that are being monitored. The screen shot is from one of the servers that do not have the SQL services on it.
I pointed out the applications that are OK to be "not running" on two of the servers, but must be on the third. I know it's probably the lazy way, but I figured I could just put logic in the alert so that it would alert on the SQL server if the services were not running, while ignoring it on the other two.
I would really appreciate any comments or ways I could have done this easier or made the alert better.
Thanks all!!