mq_coll

README

MQ Exporter for Collectd monitoring

This README should be read in conjunction with the repository-wide README that covers features common to all of the collectors in this repository.

This directory contains the code for a monitoring solution that sends queue manager data to the collectd system. It also contains configuration files to run the monitor program

The monitor collects metrics published by an MQ queue manager. The monitor program prints these metrics to stdout, in a format that is recognised by collectd. Once processed by collectd, they may be forwarded to storage systems, where they can then be queried directly or used by other packages such as Grafana.

You can see data such as disk or CPU usage, queue depths, and MQI call counts.

Configuring MQ

No MQ configuration is required, as (unlike the other monitors in this repository) the program does not run as an MQ Service.

Configuring collectd

There are several steps needed to configure collectd. It will use the 'Exec' interface to manage the MQ collection.

• Edit the mq.conf file to point at the shell script that starts the real program. The name of the queue manager to be monitored is also given here.
• Edit the shell script mq_coll.sh to set additional parameters for the program, and ensure the program is being invoked from the correct directory.
• Put mq.conf in the /etc/collectd.d directory. That is read during startup, telling collectd how to use MQ. It also contains a reference to the list of metrics generated by the program, which is added to the global TypesDB configuration.
• Restart collectd to pick up new configuration. For example

    systemctl restart collectd

Metrics

Once the monitor program has been started, you will see metrics being available. The metric names sent to collectd start with "qmgr" followed by the queue manager name. The next element is the actual metric (queue_depth, system_cpu etc) followed by the queue name if appropriate. You can check that the mq_coll program is running with the ps command. That shows correct configuration of the collectd plugin and the mq_coll.sh script.

There is some minor rewriting of object names, replacing "." with "" to conform with the collectd Exec interface; the metric names may get rewritten again by the conversion to the backend storage (re-replacing "" with "." perhaps).

The example Grafana dashboard shows how queries can be constructed to extract data about specific queues or the queue manager.

More information on the metrics collected through the publish/subscribe interface can be found in the MQ KnowledgeCenter with further description in an MQDev blog entry

The metrics stored in the database are named after the descriptions that you can see when running the amqsrua sample program, but with some minor modifications to match a more useful style.