Problem URGENT after my server goes down

Yesterday we had an energy problem and my no-break is broke, so my AS400 gone down abruptaly. After energy came back, my AS400 started in a recovery mode and took 3 hours reorganizing table indexes. After that I have the following situation:
Some UBEs that are started on server, uses CPU time and let other UBEs in a MTXW status. Seeing in OneWorld at submitted reports this UBE shows that is in waiting status and it never ends. After I end the job the others UBEs are released till another execution brings the same problem. This is happening in R42800 and about 6 UBEs and not for all the selections. Aparently only has problems with documents that was processing when the machine gone down. Other strange thing is that if I run the problematic process locally it finishes OK.
Anyone had a problem similar? Any Idea what can I do?

Thank You

Sounds like corrupted specs. Consider redeploying last full server package.

Have you tried deleting the SQL Packages? (All but the Q* and _* ones). In
some cases the AS/400 is too smart and can rebuild even damaged SQL
packages after they've been deleted so I would normally reboot the AS/400
after deleting the packages. Since you've already rebuilt the indexes the
reboot should not take too long. It's just a precaution in case deleting
the packages alone doesn't work. Everyone's experience is different.

I know there is a Q* SQL package that can be deleted. I'm just talking
about rule of thumb for now.

Regards,
Gerald




rcandeo
<rcandeo@araucodo
brasil.com.br> To
Sent by: [email protected]
jdelist-bounces@j cc
delist.com
Subject
Problem URGENT after my server goes
11/01/2006 10:46 down
AM


Please respond to
JD Edwards®
EnterpriseOne
<jdelist@jdelist.
com>






Yesterday we had an energy problem and my no-break is broke, so my AS400
gone down abruptaly. After energy came back, my AS400 started in a recovery
mode and took 3 hours reorganizing table indexes. After that I have the
following situation:
Some UBEs that are started on server, uses CPU time and let other UBEs in a
MTXW status. Seeing in OneWorld at submitted reports this UBE shows that is
in waiting status and it never ends. After I end the job the others UBEs
are released till another execution brings the same problem. This is
happening in R42800 and about 6 UBEs and not for all the selections.
Aparently only has problems with documents that was processing when the
machine gone down. Other strange thing is that if I run the problematic
process locally it finishes OK.
Anyone had a problem similar? Any Idea what can I do?

Thank You


Robson Candêo Placas do Paraná - Brazil Xe, SP20, Update7, AS400.

If my specs where corrupted, it wouldn't work for some selections, don't you think?

I have an automatic deletion of my SQL packages after an automatic IPL during the night and it didn't solve my problem. Any other idea please?

[ QUOTE ]
Sounds like corrupted specs. Consider redeploying last full server package.

[/ QUOTE ]

I agree. If UBE's were running while the system went down, those specifications became corrupted. When you restart services, the UBE's started running again with the bad specs. We just had this problem today. We redeployed packages for both DV and PD and that corrected the issue of JDENET kernals going to a MTX state.

Seen something like this before.......

1. ENDNET, CLRIPC, ENDHOSTSVR *ALL, ENDDBS QUSRWRK *IMMED
2. Delete all SQL packages except those that start with 'Q'
3. Delete global tables and data dictionary of IFS for EVERY path code while JDE is down.
4. Restart everything

Hello Robson, I've seen this type of issue and the quick solution was to remove the E1 SQL packages and re-deploy the last full package to the server. Also ensure that E1 services are coming up clear. If they are not coming up clear, then it could be an issue with the kernel spec files jdekernl.ddb jdekernl.xdb.

Boa sorte e um abraço.

Thank You for everyone that helped me. After redeploying my full server package, it all worked normally.