MTC-A Msg Server error: No OPS instances detected after mounting newly defined datasets

Products

• OPS/MVS Event Management & Automation

Issue/Introduction

We did have everything working. However, we opted to go with the default message server file size of 20GB and have recently decided to scale that back to 5GB. 

We recreated new mount point datasets at 5GB using a new dataset naming syntax (virtually recreating a fresh deploy of the message server). 

We unmounted the original mount points and remounted using the newly allocated zfs datasets. 

Upon trying to start ZMSSTART, we are receiving:

   run(adapter.java:515): No OPS instances detected

I stopped the task and unmounted the new mountpoints then remounted the old ones, all is well again. It appears as if there is an OPSMVS hook into the existing mounts that is preventing it from reading the new mounts?

 

 

Environment

Release : 13.5

Component : MAINFRAME TEAM CENTER AUTOMATION

Resolution

MTC-A does not use hooks.   

One way to use the smaller dataset is as follows:

Replace the broken adapter in the smaller mount point with the working adapter.  

1. Copy the adapter out of working version

'cp {message service path}/messageService/data/adapters/ops-admin-adapter.jar /some/uss/path'

2. Remount the smaller, non-working dataset

3. Save off the broken version of the adapter in case this doesn't work.

'mv {message service path}/messageService/data/adapters/ops-admin-adapter.jar {message service path}/messageService/data/adapters/ops-admin-adapter.jar.newVersion'

4. Copy in the working version of the adapter

'cp /some/uss/path/ops-admin-adapter.jar {message service path}/messageService/data/adapters/ops-admin-adapter.jar'

5. Recycle ZMSSTART