As our automated processes become more and more critical, it is imperative to be able to monitor those processes to know if they completed, failed, or are stuck in a blocked state. I'd like to know how everyone is monitoring their processes. How do you know if an operator has changed to the Blocked state and caused your process to stall?
Hi James,
We are are a Mainframe shop, and we have CA Workload Automation (ESP) running on our Mainframe.
I have created a Powershell wrapper that we call directly from ESP to launch our jobs on the CAPA server instead of the built-in CAPA scheduler.
This way we have the jobs monitored by our 24/7 command center and I don't really have to worry about it.
However, for one of my processes (Unix server patching) I actually wrote a C# service that runs on the CAPA server and monitors the patching jobs against a list of 'known to be active' jobs contained in a dataset.
If the dataset shows that the job is active, the service then does a web-service call to CAPA to validate the actual state of the job, and updates the job status accordingly.
That way I can catch if a job aborted (or was cancelled) as I wasn't able to trap some of those errors within the error handling routines of CAPA itself.
We are also using Microsoft SCOM to monitor the state of the CAPA services, and that the webservices are responding.
Not sure if that helps you out, but thought I would add my $0.05
(Up from $0.02 since we don't have pennies in Canada anymore)
Ian
Hi Ian,
Thanks, that is very good information. Does your Powershell wrapper poll the processes it initiates to return the process result (e.g. completed, failure, blocked) to ESP? Would you be willing/able to share the script?
Cheers!
Hi Jeremy,
Actually the Powershell wrapper runs as long as the process is active, and exits when the process completes (I'd have to check on blocked state, but I know for sure on success/failure/aborts). It also runs the process with the --trace switch so that all spool output is actually captured to a uniquely generated log file. That log file is then parsed for specific variable keywords (ESPExitStatusCode, ESPExitReason) which are then passed back to the calling ESP job, and then deleted.
If you are interesting in seeing the script, I could take some time to remove all our system-specific information in it and post it for you. It is fairly well documented (for a script) but I could also try to explain how it works a little bit .
Thanks,
Ian
Using ESP for your scheduled processes is a great idea. Since we also have ESP installed, it's an idea I will certainly entertain as we begin to amass critical scheduled processes. Seeing your PowerShell script would be very beneficial to me when we decide to pursue this course of action.
Hi Guys,
Okay I did my best to clean up the scripts a little bit, and I wrote up some general instructions.
Hope you find it useful!
https://communities.ca.com/docs/DOC-231151576
Ian
Hi James, I usually use in my process the Exception Handler, with it I control my process to do any else if a problem occurs. Some people use the default report to do it too. I know, but I never used, the PAM has resource to monitor itself, but you need to do a process to do this, searching for the processes and looking the status of they.
Let's see how other people do it.
If I'm wrong, please someone correct me please.
Best regards