I would like to monitor the "C/S Requests Pending" attribute of 60 POAs every minute without GroupWise Monitor.
It happens from time to time, that C/S Handler Threads are maxed out (99) on random POAs and users complain about login problems, slow client responses. After 10-15 minutes the threads are freed up without any intervention. I would like to investigate this problem when it is in progress, so I need almost real time data and notifying about this. There is a working Icinga system with a well proven action plan and I would like to have a service with the C/S Requests Pending or C/S Handler Threads in it.
I do have GW Monitor but the notifying capabilities are limited compared to that if Icinga.
I can download the web interface (port 7181 )with wget and grep for the required strings but this looks like an overkill for me. I need only one bit of information but I download the whole web page. And on the other side I suspect that the web interface is serviced with the C/S Handler Threads, so if all threads are busy, I cannot reach the web interface either. GW Monitor is not affected by this, it can read the numbers.
I wish I could use the same method as GW Monitor, but all I've found is that it is polling the POAs over XML. Whatever this means.
How does GW Monitor polls the GW agents, and how can I use this method to have my own monitoring in Icinga? I assume it is proprietary and I cannot use it, but if I don't ask, I don't get answers :)