When configuring SiteScope MQ monitor, you need to select queues from a list, in which there are local queues and remote or alias queues. An then select counters about you want to get information.
From the MQ standard definition, you only can get information from local queues, not from remotes or alias queues. From an efficient point of view, I think it’s not the best option to query all MQ queues when you only can get information from local queues. Monitor will last much more time to run. In my env, now we are testing with a MQ manager with “only” 400 queues, but there are other MQ managers with 3000 – 4000 queues… It’s quite inefficient to query for remote queues when we aren’t going to get the relevant data and process the same. A proactive approach would be not to query for remote queues information in the first place. And, this will result in a memory, network, & compute time efficient solution.
In a customer point of view, it’s more acceptable to take time displaying a list of only local queues than trying to get no information from remote/alias queues. Also, displaying a list of local queues is done only one time, but trying to get information from all (local and remotes) queues is done every X minutes (according to monitor settings).
An idea about how MQ monitor should work:
- The SiS monitor query to MQ manager should search for attribute ‘TYPE’ and get only those with value ‘QLOCAL’. For example:
- And the SiS monitor query to MQ manager should discard ‘QALIAS’ or ‘QREMOTE’ values. For example: