I only had similar problems once in 2008 ca., but then subtasking was
involved
and APL etc., and there was some sort of race condition where one subtask
ran too fast and the POST/WAIT protocol dit not work ... ABEND S201,
IIRC ...
the error was in IBM code, not in ours.
The problem occured only on newer hardware with many processors, only every
5 hours or so, and only when there was heavy load by other concurrent jobs.
IBM maintenance was not able to reproduce the error, because our machines
were faster than theirs.
Very hard to find ...
I don't think that such problems may be involved here ...
another idea that comes to mind: areas of storage that are normally zero,
but now due to opsys release changes etc. are non-zero, and the application
(which by error doesn't initialize some automatic variables properly)
now fails.
That is: a programming error that was there all the time, but now shows up
for the first time.
Don't know, if those "war stories" help in any way ... without looking
at the
dumps, I can't do any better.
Kind regards
Bernd
Am 10.10.2014 08:05, schrieb Chase, John:
It appears that we captured relevant problem data via GTF tonight, so "we shall see
what we shall see, when we see it".
-jc-
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
