help! ubes take a very long time to go to processing status

[msjxx]

when a job is submitted to the enterprise server, it takes a very long time for the job to go from waiting to processing even though there are queues available. this can be up to a few mins. in db2, we see a high wait time for sql processing against the svm7334.f986110, but if i query the table from a command line, i get an immediate response.

does anyone have any ideas?

platform:
erp 8.0 sp22_f1
db2 v7.1
aix

 

[JasonPenner]

Can you post the part of your client side JDE.INI file from the [NETWORK QUEUE SETTINGS] section?

 

[msjxx]

[NETWORK QUEUE SETTINGS]
UBE Semaphore Key=3600
DefaultPrinterOUTQ=oneworld_a
UBEQueues=15
UBEQueue=QB7334
UBEQueue=QB7334
UBEQueue=QB7334
UBEQueue=QB7334
UBEQueue=QB7334
UBEQueue=QB7334
UBEQueue=QB7334
UBEQueue=SINGL001
SpecInstallQueues=1
SpcQueue1=QB7334

 

[satishjk]

Add this to the network queue settings
QueueDelay=5

You might have to jump the services after this change.

 

[msjxx]

in the runque.sh script, the queuedelay is already set to 5.

 

[verberr]

We have run into this as well. We have reanalyzed the svm7333.f986110 table and that took care of the issue for us.

Hope it helps

 

[msjxx]

we actually re-index and ran statistics on that table but did not help.

 

[mkkanth]

Two things to check:

1. Externally Run the select statement on the F986110 table where jcjobsts='W' (with queue and without queue) and see how long this one takes to come back. Sometimes database chooses not so good index to do a query than what we think it should.

2. Double check the StartOneWorld.sh script and make sure the queue's interval are set to very minimum in the startup script.

These are the two things we ran into when we had similar issues.