Memory error when job scheduled

[YaleM4208]

We just migrated a bunch of jobs to our new 11.3 environment. In one sequence job, we have a Routine Activity stage that calls a .bat script through the built-in ExecDOS routine. The .bat script calls the MKSToolkit zip command and moves some files around. There are 8 files being zipped, totaling about 225MB. The script runs fine via a command prompt, and the job runs fine when run directly from the Director or Designer client. However, when the job is scheduled through the Director and run automatically, the zip command errors - "zip error: Out of memory (allocating temp filename)" - and the sequence aborts.

Is there some way to adjust the memory available to automated jobs?

[ArndW]

The difference between your run methods is most likely attributable to the different user-id and permissions (and perhaps environment) when run via batch versus your other call methods and not necessarily due to actual memory constraints for the zip command.

It might be file access permissions in the directories that are missing for the batch user.

[YaleM4208]

We changed the script to use the 7-Zip 7z command, rather than the MKS zip command, to create the archive and initial testing looks like this resolves the issue. I opened a PMR with IBM for the initial error though and will post back here if they provide any useful information.

[chulett]

I was under the impression that MKS was out of the picture in the newer versions so perhaps the zip it was using is just what is built into Windows, the so-called 'compressed folders'?