Re: SQL EXECUTE failing after upgrade to V6R1

David FOXWELL wrote:

rob@xxxxxxxxx wrote:

Since IBM is able to reproduce this, I'll think I'll step out
of it...

I just sent the post as a warning to others, I'm just surprised that I've not read of a similar problem on the list. We seem to
be later than others in passing to V6R1. I was hoping though,
that someone might say that the PREPARE/EXECUTE wasn't being
coded correctly. It's not my code and I've never used SQL in
that manner before.

FWiW the problem was presented too generically and has too few problem-determination details provided. The error being seen is a catch-all, the SQLCODE -901 [SQL0901 which even has a return code & previous message identifier when extracted from a joblog; symptom msgSQL0901], which is an indication that the OS database SQL feature suffered an "internal failure". Often the error is preceded by other messages in the joblog, often "machine check" errors; minimally, those errors should be included for someone to try to comment on the SQL0901 error.

The -901 is generally an indication of a defect in the SQL feature itself, and thus is seen irrespective of the [im]proper coding of the SQL by the programmer, and typically also seen irrespective of the when, where, or how the program is called. There is nothing strikingly wrong with the code as described; nor even as posted, as inferred from a very cursory review.

Unless actual source with compile instructions for the program and referenced objects plus a command string to invoke the program could be provided to recreate the error on another system, those outside of the IBM Rochester database SQL developers are probably not the best people to offer some advise other than a PTF that might record the symptom for the failing statement. IBM support seems the proper place for the subject issue, and hopefully you will be provided with timely resolution or a circumvention.

Regards, Chuck