Qzdasoinit jobs on AS/400

Toni_Nanneman

Active Member
We have 750 prestart Qzdasoinit jobs starting for OneWorld. Lately, we've
been having slow down issues and it appears that new ODBC connections are
being started which would account for the possible slow-down times. We are
worried that the old odbc connections are not being 'freed' when the user(s)
close out of an application or connection. We recently upped our pre-start
jobs on the 400 from 500 to 750 and continue to have this problem.

For example: today we had 35 users logged on and the dspactpj entry shows
that there were 800 active prestart jobs! No WAY!

Any ideas or helpful hints?


Toni Nanneman
CNC Administrator
503-370-7071 x7262
[email protected]




B7322 SP12.2; NT; SQL 6.5; coexist on AS/400 with A73 cum 8
 
Toni :

You should migrate to SQL 7 as soon as possible.
In the mean time, be sure to have the latest SP of SQL 6.5 applied (SP5a).

A lot of luck, Sebastian

-----Mensaje original-----
De: Toni Nanneman [mailto:[email protected]]
Enviado el: Thursday, March 08, 2001 8:55 PM
Para: [email protected]
Asunto: Qzdasoinit jobs on AS/400


We have 750 prestart Qzdasoinit jobs starting for OneWorld. Lately, we've
been having slow down issues and it appears that new ODBC connections are
being started which would account for the possible slow-down times. We are
worried that the old odbc connections are not being 'freed' when the user(s)
close out of an application or connection. We recently upped our pre-start
jobs on the 400 from 500 to 750 and continue to have this problem.

For example: today we had 35 users logged on and the dspactpj entry shows
that there were 800 active prestart jobs! No WAY!

Any ideas or helpful hints?


Toni Nanneman
CNC Administrator
503-370-7071 x7262
[email protected]




B7322 SP12.2; NT; SQL 6.5; coexist on AS/400 with A73 cum 8
--------------------------
To view this thread, visit the JDEList forum at:
http://198.144.193.139/cgi-bin/wwwthreads/showflat.pl?Cat=0&Board=OW&Number=
6978
*************************************************************
This is the JDEList One World / XE Mailing List.
Archives and information on how to SUBSCRIBE, and
UNSUBSCRIBE can be found at http://www.JDELIST.com
*************************************************************



B7321 to Xe, NT/W2K/SQL
JAS, Interoperability
MCDBA,MCP+I,MCSE,Citrix Admin
[email protected]
Grupo ASSA - Application Software SA
 
It would appear that SSAJAROFF is missing something here. This has nothing
to do with SQLServer.

There are several things to looks at. First look at the Prestart Job entry
QZDASOINIT in the QSERVER subsystem description. This will tell you how
many get automatically started when the subsystem comes up. IF you don't
take down QSERVER very often, these jobs may increase and never end. They
will stay active until the subsystem comes down or the Max Number of Uses is
reached (and the OS will end them automatically). How often do you take
down QSERVER? Next, try doing the DSPACTPJ QSERVER QIWS/QZDASOINIT. There
will be a number in use and a number of jobs active. If the number of jobs
in use is relatively low but you have a large number of jobs running, you
are probably ok. Each OneWorld session uses 8 database connections
(QZDASOINIT). When they are not in use, they don't really affect
performance much. What you want to look at is the prestart jobs in use.

I have found OneWorld database connection performance to have less to do
with these ODBC connections and more to do with the number of JDENET and
Security kernels that are active in the subsystem.


>From: SSAJAROFF <[email protected]>
>Reply-To: [email protected]
>To: [email protected]
>Subject: RE: Qzdasoinit jobs on AS/400
>Date: Thu, 8 Mar 2001 16:24:56 -0800 (PST)
>
>Toni :
>
>You should migrate to SQL 7 as soon as possible.
>In the mean time, be sure to have the latest SP of SQL 6.5 applied (SP5a).
>
>A lot of luck, Sebastian
>
>-----Mensaje original-----
>De: Toni Nanneman [mailto:[email protected]]
>Enviado el: Thursday, March 08, 2001 8:55 PM
>Para: [email protected]
>Asunto: Qzdasoinit jobs on AS/400
>
>
>We have 750 prestart Qzdasoinit jobs starting for OneWorld. Lately, we've
>been having slow down issues and it appears that new ODBC connections are
>being started which would account for the possible slow-down times. We are
>worried that the old odbc connections are not being 'freed' when the
>user(s)
>close out of an application or connection. We recently upped our pre-start
>jobs on the 400 from 500 to 750 and continue to have this problem.
>
>For example: today we had 35 users logged on and the dspactpj entry shows
>that there were 800 active prestart jobs! No WAY!
>
>Any ideas or helpful hints?
>
>
>Toni Nanneman
>CNC Administrator
>503-370-7071 x7262
>[email protected]
>
>
>
>
>B7322 SP12.2; NT; SQL 6.5; coexist on AS/400 with A73 cum 8
>--------------------------
>To view this thread, visit the JDEList forum at:
>http://198.144.193.139/cgi-bin/wwwthreads/showflat.pl?Cat=0&Board=OW&Number=
>6978
>*************************************************************
>This is the JDEList One World / XE Mailing List.
>Archives and information on how to SUBSCRIBE, and
>UNSUBSCRIBE can be found at http://www.JDELIST.com
>*************************************************************
>
>
>
>B7321 to Xe, NT/W2K/SQL
>JAS, Interoperability
>MCDBA,MCP+I,MCSE,Citrix Admin
>[email protected]
>Grupo ASSA - Application Software SA
>--------------------------
>To view this thread, visit the JDEList forum at:
>http://198.144.193.139/cgi-bin/wwwthreads/showflat.pl?Cat=0&Board=OW&Number=6979
>*************************************************************
>This is the JDEList One World / XE Mailing List.
>Archives and information on how to SUBSCRIBE, and
>UNSUBSCRIBE can be found at http://www.JDELIST.com
>*************************************************************
>

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com



JD Nowell
OW: B7332
ES: AS400 V4R4 CO: DB2/400 SP: 11.2
Users: 250 TSE Users: 100
 
800 prestart jobs is between 100 and 200 users (depending on how usual your
OCM/data sources are set up). You could check the max number of re-uses.
If set to 1, each new connection requires a prestart to be initiated. Here
are a couple of other ideas:

You should check the size of temporary space on wrksyssts. As temporary
space allocation approaches the total size of disk, the machine can bog
down.

If your users start to use business views that use the union verb over large
tables then perform sequencing using columns from both tables, the machine
will work pretty hard at that.

Richard Jackson
 
Please clarify "relatively low." We have 720 prestart jobs. I've been
watching the current number in use throughout the day today and it's been
between 400 and 500 each time I've looked. Although, it's reporting the
average number to be 291.2 and peak is 533.

We restart QSERVER every night.

Is this good? Bad? Other?

Michelle Dulay
ERP Coordinator
BFGoodrich Aerospace
Chandler Evans Control Systems
email: [email protected]

OneWorld B7322 with the MFG Mini-Cum, coexistent with World A7.3
AS400 V4R3 Enterprise Server
Windows NT SQL V6.5 Deployment Server
SynQuest V5.07.02
Oracle
NT Logic - between AS400 (OneWorld) and HP9000 (SynQuest)
 
Re: RE: Qzdasoinit jobs on AS/400

"Relative" would refer to the average in use as a percentage of max active. "Relatively low" would be an average prestart jobs in use of about 25% of prestart jobs active. "Relatively high" would be an average prestart jobs in use that is 90% of prestart jobs active.

JD Nowell
OW: B7332
ES: AS400 V4R4 CO: DB2/400 SP: 11.2
Users: 250 TSE Users: 100
 
Back
Top