Checking status of running task
I have been working on a project in our Dev environment which resulted in processing a few (million) HL7 messages, and I thought it would be a good idea to purge the namespace in a controlled way. My concern was when it hit the day where the scheduled task would purge 'the day of a million messages', it would fill the Journal Drive quite dramatically and would happen overnight resulting in triggering a response from the poor person on out of hours support.
My idea of a controlled way was to:
- Run a backup
- Disable Journalling
- Run Purge
- Compact and Truncate DB
- Re-enable Journalling
Now, I got as far as running the Purge and (surprise surprise) it's taking a little while.
I was wondering if there was a clever way of checking the progress of the task (beyond seeing that it is running)?
On top of this, can anyone suggest a different method or approach to the above that may have worked better?
Like I said, this is a Dev environment on it's on dedicated server so I'm not too worried. But it's better to learn from a mistake in Dev than in Live
Comments
I love Docker way. In Dev environment, when you need only some test data. You can use Docker, with all test data inside the image, just recreate container will clean all previous changes and return the database to state when this image was built. If you later will decide to keep some changes as default for next uses, you can commit this changes to the image.
Docker sounds interesting, I'll give it a try next time.
Thank you - I will try this the next time I'm running a big job and see what details I get.
Having a refreshable image of your Cache/Namespace test area is the way to go.
Use Docker, VM machine image, backup image or other techniques to quickly reset full Cache image back to baseline for more testing.
Tom Fitzgibbon | gototomAtG...l | 13474648531
For checking the progress of deletion of older messages, just bring up the message viewer and set the Sort Order to Oldest First and the Time Format to Complete. Each time you click the Search button, you should see progressively "newer" messages as the older ones are purged.
You can use ^JOBEXAM in terminal or the process details page in the System Management Portal to see the current state of the job. This gives you access to the process's current variables which should include the current row ID being purged.