Thanks for trying to help Torsten and Gerard!
It appeared to be something else and I now know what the problem was. I'll try to explain.
I found out that the message queue (in general, not the sensor) only works with the real server hostname and not with alias names.
When we get our servers deliverd, the servers get an hostname like PL390.
Because we have a lot of servers, we give all the servers an alias like appserver-prd01 for example.
Because that is the name we use for the servers, I've always been monitoring the servers by their aliases. It works perfectly for services monitoring and custom wmi scripts and I didn't know that the message queue only works with the real server name, so I just added the sensor to the devices without changing the servernames PRTG was looking at.
Anyway, I had to change the IPv4 Address/DNS Name for all my servers and now the Windows MSMQ Queue Length sensor works again.
The only thing I don't know now, is how make PRTG give a down-alarm when there are like 5 messages in the queue and a warning if there are two. I do get to see the values now, but without the ability to give up a 'Condition Warning' and a 'Condition Down', I don't really see the point monitoring the amount of messages.
Where should define those conditions?
Add comment