Could it be possible the RPG UDF is being called recursively? The
program developer may have to modify it to ensure it is closed (*INLR =
*On) after every use. If it needs to be called recursively changing its
activation group to *NEW will allow recursive calling. An activation
group of *NEW may not be the most efficient method but it works.
-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of DebbieKelemen
Sent: Friday, March 28, 2008 8:03 AM
To: 'Midrange Systems Technical Discussion'
Subject: RE: QZDASOINIT / SQL - Stored Procedure Question
Dan,
I know they are there. It's when a SQL Stored Procedure won't run again
without getting an error because it is or has already been run.
Actually, the big one we have an issue with is an RPG program with is
set up as a UDF which has an issue if it is called by the same person
more than once. I have tried setting this up multiple ways but get the
same results.
I'm stumped with what they are doing to cause this.
Deb
-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
dkimmel@xxxxxxxxxxx
Sent: Thursday, March 27, 2008 5:21 PM
To: midrange-l@xxxxxxxxxxxx
Subject: RE: QZDASOINIT / SQL - Stored Procedure Question
You don't need to kill them off. Those jobs should always be there to
handle new requests. The jobs do not end when a connection is closed,
they stay active in a wait state to handle the next request. These jobs
are prestart jobs; there will be three or four of them active as soon as
the subsystem is started before any connecitons are made. Each one
usually handles a hundred requests or so before they are eligible to
end.
If you kill them all off, three or four new ones will start. The number
that prestart is configurable in the QUSRWRK subsystem. Also the number
of uses and several other parameters.
The jobs don't actually stay there forever, its just that whenever one
ends, a new one (or several more) start up.
Dan
-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of DebbieKelemen
Sent: Thursday, March 27, 2008 5:06 PM
To: 'Midrange Systems Technical Discussion'
Subject: QZDASOINIT / SQL - Stored Procedure Question
I have searched the Midrange-L archives and I realize this topic has
been discussed a lot. I couldn't find a definitive answer. So I
thought I would see if someone can give me a quick & easy answer.
We are on V5R4. Our QZDASOINIT jobs run in QUSRWRK which has it's own
storage pool. My issue is these jobs stay in the system forever (or at
least until our next IPL) or so it seems.
The problem with this is our web folks have some SQL in Stored
Procedures which they run to extract data from the iSeries to build
other tables on the iSeries. Occasionally, when they are testing these,
they run them from WDSC. When they do this it seems to leave things
"hung" in one of the QZDASOINIT jobs. Even if the developer closed his
pc down & reboots. If he tries to rerun the process, it gets an error.
If I go manually end a bunch of the QZDASOINIT jobs, and I have to be
careful here because our website jobs come in through these jobs also,
then they can rerun their process.
I have often thought they are closing things correctly in their SQL
processes but I'm not fluent enough in SQL to know. After reading the
issues with the QZDASOINIT jobs hanging, is it something else?
Is there something I can be doing on the iSeries side to clean these up?
Debbie Kelemen
Sr. Programmer / Analyst
chefs
719-272-2617
www.chefscatalog.com
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing
list To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe,
unsubscribe, or change list options,
visit:
http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take a
moment to review the archives at
http://archive.midrange.com/midrange-l.
midrange.com
