When a server agent goes offline, we have an engineer on duty that responds to the alert notification 24/7. Basically, so the engineer doesn't have to be constantly looking at their email, we have a TAPI dialer hooked up to our server that is triggered by a command line script. This executes a more complex series of actions involving a phone tree and an alert that must be cleared by the engineer so it doesn't go down the chain in our phone system.
So in Monitoring > Alerts > Agent Status we configured a script that executes when an agent goes offline. Additionally, an email is set to go out to our entire team notifying everyone that a server is offline.
What I would love to do is setup an ONLINE notification that tells the team that the server is back online. However, in order to do that, I have to check the "Alert when agent goes online" checkbox. What this does is needlessly trigger the script action AGAIN and annoy the engineer on duty because the whole phone tree is triggered again, and they must clear the alert in our phone system. I'd love to have the agent ONLINE action execute without the script so this doesn't happen.
Does anyone have any ideas for accomplishing this?
Hmmm... This is a really interesting question. I would open a ticket with Kaseya to see if there is a variable you can reference in your script that will tell you what the alert being generated is. If you can kick out the alert from a variable, you can use an IF statement to parse for "online" and simply exit the script if you find it.
Sorry I can't be of more help, but if you do find the answer please post back here. I'd be interested to know!
Fire a script to run when the agent goes down, do not set it to skip if the machine is offline, that way it'll sit in queue until the agent comes back.
Have the script write an event log entry and email you.
Good suggestions. The complication is that I have some logic to check to see if other servers at a client site are down before making a phone call. This prevents our engineer from receiving 20 phone calls for 20 servers. The consequence of this is that I lose a handle on who the original intiator of the script is.
So I can't write a log entry to the original server - the script was triggered on another server, so I would be writing the log file against that other server.
Nagios' Alarm Grid shows when a server is offline or has gone offline with different colored icons. I have tried to convince one of the Kaseya Trainers that this is they way to go with their alarm grid monitor dashboard however no luck so far. Its sad that the wheel gets re-invented every time something gets done in the private sector.