5 Tips for Monitoring WebSphere MQ like a Pro
1. Monitoring queue manager status
This most basic monitor is the one that is often overlooked when MQ users are focused on setting up queue depth alerts. Often queue depth alerts do not return anything at all when queue manager is down so it is important to have a separate monitor just for monitoring queue manager status.
2. Stop alert floods.
For example if you have a queue manager down situation, your monitoring system should not send out all the queue depth and channel status alerts or you will get 100s of alerts overwhelming support staff and mail inboxes. Smart monitoring system should suspend all other alerts when queue manager is down
3. Leverage naming convention.
If you rely on a naming convention, then you already know your error and back-out queues. Good monitoring practice is to setup queue depth > 0 alerts for all error and back-out queues. There is not need to specify this manually. So when you create a new error queue, good monitoring system will begin to monitor it automatically.
4. One server connection channel per consumer.
This allows you to define MCA User with security permissions specific to that client and also if that channel is not running you know exactly which consumer is down.
5. Detect problems before data is lost or corrupted
If you notice that consumer is not connected, then you do not have to wait for the queue to fill up. You know it will fill because consumer has crashed and no longer consuming messages.