Heh. I think I worked on that particular machine way back when. :-))
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Chris Bipes
Sent: Monday, March 17, 2014 3:26 PM
To: 'Midrange Systems Technical Discussion'
Subject: RE: TIMA cpu use VS DLYJOB
I worked on a NEPS way back on a B10 and OS400 2.3. This was a real AS400
system. If you do not change the reuse delete to *NO, your job will wait at
the end of the file while records are written in the middle of the file.
These records will not be picked up.
I made that mistake to help with the month end processing. Job sat for a
week before someone noticed that the summaries were not being updated.
So I highly recommend you set Reuse Delete to *NO.
Director of Information Services
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of DrFranken
Sent: Monday, March 17, 2014 1:16 PM
To: Midrange Systems Technical Discussion
Subject: Re: TIMA cpu use VS DLYJOB
I used to have a screen capture from the thing but I can't find it now.
Rats. It was a truly inspiring number I think in the Trillions.
I don't recall the rule about reuse deleted records but there is some
logic to that to me. It is not in the help text for OVRDBF.
Also if you read the help text for the EOFDLY parm it doesn't, to me at
least, correctly describe the behavior. It actually says the job will
wait that long (as specified on EOFDLY) before it will attempt to read
another record. But what it actually does is get the next record written
to the file immediately. The delay says if NO records are written for
one EOFDLY interval then the read fails with EOF. Hmmm.