Requesting assistance for accessing the persistent volume of an MQ container Native HA queue manager deployed in the GCP GKE cluster when the POD fails to recover

[vmamid-cswg]

Good morning Team,

We deployed IBM MQ queue managers as containers using the Native HA approach in a GKE cluster on GCP cloud by using Helm chart provided by IBM.

To troubleshoot issues, we typically connect to the MQ queue manager POD to access the queue manager error logs.

However, for one of the queue managers, we were unable to bring it up (the POD could not start), preventing access to the error logs stored in the persistent volume.

Could you advise on how to connect to the persistent volume without a running POD to access the error logs for troubleshooting purposes?

Thanks and Regards
Ganesh Mamidi

[arthurbarr]

You can either use the mq-container-inspector tool, or follow the instructions here and use the mq-pvc-tool.

Another useful trick which can sometimes work, is to set the environment variable DEBUG=true on the container, which prints some additional information to the log.

[vmamid-cswg]

Thank you for the details.

As per the documentation, these tools are available for IBM MQ Operator on Red Hat® OpenShift® only. However, we deployed IBM MQ queue managers as containers using the Native HA approach in a GKE cluster on GCP, rather than through IBM MQ Operator.

Please advise if we have any other possible solutions. Thank you for your time.

[arthurbarr]

The mq-pvc-tool is very simple, and is just a shell script to run another Pod which mounts the volumes. It does use the oc command, but that's almost identical to kubectl, so you should be able to edit this easily. If not, you can hopefully still understand the principle of adding a new Pod on the same Node.

FYI @Andrew-Lees11 ^^