Description
Ask a question
To test my monitoring setup I tried to make a monit monitor fail to see how that would be shown in icinga.
The monitor in monit shows initializing as Status and Monitoring status in the monit webinterface. It doesn't fail completely, because monit tries to restart the process again and again.
This makes the monitor simply vanish from the output of check_monit.py. Is this behavior planned and wanted this way?
My expectation (without being experienced with monit nor icinga) would have been that if a service can't initialize completely and therefor is stuck in the status initializing that it would be interpreted as broken and therefor WARNING or CRITICAL on icinga side.
I looked at the xml generated by monit, but didn't find a list that would explain the integer numbers used for service type, status, monitor and the like.
It seems that the monitor in state initializing vanishes, because the value of monitor checked here is 2 and nothing is output.
Also service_type == 3 is not taken in account in service_output. I understood that this only affects that there is no meaningful output in the form of a text.
I changed the script a bit and now it seems to work like I expected, but this is just by looking at the xml and guessing what the meaning of the attributes and their values could be.
Could a proper handling of service_type == 3 and for status initializing be added? Could you provide a reference that helps to understand the xml?