Job in Waiting Status before running

cati

Active Member
We have just noticed that our jobs in stay in waiting status too long in all environments. They will wait anywhere 2 minutes even if no other jobs are running.
We have an e812 system tool release 8.96.1 - AS/400
The jobs all finish successfully, but we have to wait long time before the batch starts to run.
I noticed a lot of post about and we made this test
-changing a setting in the Network Queue Settings of the jde.ini file adding queuedelay=5
-reindex F986110
-delete SQL Package
-delete PDF from F986110

But nothing of this actions have an effetc.
Could someone help me
Any suggestions will be appreciated.
Attached our jde.ini
Thanks
 

Attachments

  • 162884-JDE.txt
    11.8 KB · Views: 458
I have tried to make another test putting QKActive=1 in our JDE.ini in the section [NETWORK QUEUE SETTINGS] because I have written on Oracle site that after 8.9 release the parameter QueueDelay=5 doesn't work.
But also after this change nothing change
I have no more idea which is the problem.
Could somebody help me?
 
Another thing....
The time the batch stays in waiting is 60 sec and it never change
 
Hi Cati,
I've just noticed your post. We are an iSeries site running 8.12 8.98.2.3. When the jobs are stuck on a wait status are they waiting on a queue or are they already processing from an os/400 point of view i.e.do you see them in WRKACTJOB...? If so what status are they is such as icfw, msgw, cndw etc
Rich
 
No I see ythe batch in waiting on P986110. On AS the batch goes under E812SYS/PRINTUBE while it is in waiting on F986110. Then after fixes 60 sec the batch start to run normally.
I have no idea which is the problem .
I have open a call on JDE but they don't find a solution at the moment
 
Wild guess here. Might there be "dlyjob dly(60)" in the CLP calling P986110?
 
You would look in the source file member for that CLP. Following JDE convention that member name would be have "J" instead of "P" in the first position.
 
Where are your system tables (System - 812) located in relation to the AS400 where these are being processed? Are the system tables located on the same machine that the UBEs are processing on? The UBEs do a lot of initialization and reading of the system tables before they actually start to process. For example, our Dev iSeries server takes 20-30 seconds for the UBEs to go from W to P status because our system tables are on our Production server and we know we have some constraints to fix on the NIC card of our Dev server in order to speed up that time. The log messages I put below represents the time it is doing this process. The first message "INITIALIZING RUNBATCH" is created as soon as it starts running in the subsystem, however, the job doesn't show as processing in WSJ until it gets to the last message shown, which is almost 30 seconds in this example. Almost no one uses this server so it hasn't made it on the high priority list to fix yet. Make sure the data source is not defined as JDBNET either.

226 Mon Nov 1 10:38:14.552048 runbatch.c320
INITIALIZING RUNBATCH job: 18120

226/13 MAIN_THREAD Mon Nov 1 10:38:14.623376 ipcmisc.c299
process 226 <E812SYS/PRINTUBE> registered in entry 5

226/13 MAIN_THREAD Mon Nov 1 10:38:14.630144 lanchube.c3279
Error : unlink temporary file

226/13 MAIN_THREAD Mon Nov 1 10:38:14.630440 runbatch.c465
Startup for User=USER1, Env=DV812, Role=*ALL, Job#=18120

226/13 MAIN_THREAD Mon Nov 1 10:38:32.623288 zdrv.cpp328
Initializing the Z Driver

226/13 MAIN_THREAD Mon Nov 1 10:38:32.623504 zdrv.cpp342
Gettting user data from hEnv

226/13 MAIN_THREAD Mon Nov 1 10:38:32.623592 zdrv.cpp364
User data is USER1, DV812

226/13 MAIN_THREAD Mon Nov 1 10:38:32.623776 zdrv.cpp373
Creating XML Session

226/13 MAIN_THREAD Mon Nov 1 10:38:42.403776 runbatch.c1157
RUNBATCH: Remote CP=1252, Remote OS=5, Local CP=37, ConvertToASCII=0
 
We have the table on the same enviroment where the batch runs. I don't undestand which verify I can do.
What the mean of "Make sure the data source is not defined as JDBNET either".
Where can I check that? on data source definition?
We open a call on oracle site and the answered to delete this three lines on jde.ini
QKActive=1
QKOnIdle=15
QueueDelay=5
We did these change but nothing happened.
I don't know what is but I'm sure the solution it's on jde.ini or some set-up on queue on AS/400
 
Another important thing I find this morning
The customer has an application server NT in this application the batch starts wickly.
Anybody has an idea?
 
maybe you need more ube / call obj kernels...? the F986110 doesn't always update immediately but the ube jobs should start straight away. Are they not submitting until 5 minutes or is it just that the file isn't updating quickly
 
Back
Top