Hallelujah and hosanna, Re: Erratic behavior in trigger program, since unscheduled shutdown last night.

Elvis Budimlic wrote:

- Have you signed off/signed back on since you've recompiled your trigger
program? Perhaps DB is working with the copy out of QRPLOBJ.

I just did that, right before reading your message.

I tried recompiling the trigger program with a TIME statement positioned to be executed immediately. And it threw the CPF1850 immediately. With a structurally-identical file in a completely different library.

Then I tried reworking the test program as a rudimentary trigger, but before recompiling it, I executed it once again from the command line. Wonder of wonders, it was throwing the exception, where before, it had steadfastly refused to do so. So I didn't bother recompiling that one.

So I thought, well, maybe there's something in the job that was corrupted by whatever was causing the exception in the first place (whether it was Elvis's need for an extra character on the receiver, or the inherited coding error Bruce spotted), that would be reset if I signed off and signed back on.

So I signed off and back on. The test program was back to steadfastly refusing to throw the exception. Tripping the trigger in the alternate library also refused to throw the exception. So I reapplied the trigger to the file where the problem had first shown up (I'd had to remove it because somebody else needed the file), and tripped it there.

STILL DIDN'T THROW THE EXCEPTION! BY GOLLY *SOMETHING* MUST HAVE WORKED AFTER ALL!

Since I have no way of knowing whose suggestion(s) solved the problem, I'd like to thank EVERYBODY who responded.

Somehow, the prospect of trying to get technical support for a V4R4 box, or having to bump it up a whole bunch of release levels at once if the OS was screwed up and we couldn't find anybody willing and able to help us unscrew it, didn't exactly appeal to me.

--
JHHL