DSXchange

Yes, transformers invoke the C++ compiler and if you have a single concurrent license, it could be "slow" as you wait for the compiler to come free.

Upgrades should be painless, they were before 8.x but no clue in our new world. And no need for a recompile, it handles all that automagically.

Of course, that's a given. Do you understand why? Can you answer the other questions, please?

Same way - by using the appropriate date format string: %eee off the top of my head. Check the docs.

Display? And by "day" do you just mean the day number (1-31) or the day name? Regardless, check the DateToString() function and use the appropriate 'Date format string'.

Or get the limit raised. A 2GB limit is so... 80s.

Yes, that would be best. A script can also be used to pass back a non-zero return / exit code and the job will recognize it, however (from memory) it will log a warning rather than abort.

Please, for goodness sakes, answer my dang questions regarding your job design. If your design is wrong, we could be chasing wild geese that a simple redesign would solve.

I would guess so as well. Does this happen if you run the job on a single node?

Don't think he meant that one.

How is this any kind of WORKAROUND? You are (or certainly should be) Resolved now.

Exactly, it's an average.

Well, there you go. Read all about RCP in the Parallel Job Developer Guide, page 26. You are missing a critical piece of information.

For Multi-Instance jobs, the JobName would be the base job name, followed by a dot and the Invocation ID. Then get a handle for that.

Please use meaningful subject lines, it helps searchers of the site. Your Subject and Additional Info should have been swapped, that or just skipped the whole "Hi" part as we're quite the friendly bunch and are all about the solutions. :wink: All that can be done "after job" from...