Update problem - VBDATA table

[Former Member]

Dear All,

We have update problems in our Quality systems, and also in every R/3 system that we create from a Production system copy.

Sometime the UPD and UP2 work processes hung without apparent reason.

In the SM50 I can see one or more of my update processes running for a lot of time (many hours) on the table VBDATA without perform any action.

In the tcode SM12 I donu2019t see any kind of lock, and I donu2019t understand why only sometime this problem happens.

Of course, from the SM13 I can see that during this problem the system is not able to perform any kind of update.

As I told in the beginning of the message, I have the same problem in other two R/3 systems, both created with a system copy from the production system.

The Production and the Developer systems seem not affected by this problem.

Have you got any idea about the reason of the problem?

Even if the Production systems is not directly affected, all the IT department is not able to work on the Quality system (or any other Test system), and the only workaround is to kill the UPD (or UP2) WPs affected.

Thanks in advance,

Federico Biavati

My R/3 System is a 4.7 on SLES9 and Oracle 10.2.0.2

[stefan_koehler · Accepted solution]

Hello Federico,

it seems like there is a lock on the oracle database.

Please execute the transaction DB01 and ST04 (-> Detailed Anaylsis -> Oracle Sessions) and post the output here, if the problem comes up again.

Maybe if you know howto use the AWR, you can create a report in the timeframe - so we can take a look at it right now.

Regards

Stefan

[Former Member · Accepted solution]

Hello Stefan,

No locks appear during my problem, and the Oracle session is different every time).

Iu2019ll start to investigate with AWR.

Thank you for your help.

Federico

[sivaramakrishn]

Hi,

VBDATA is a temporary table to hold update data. So, no reason why the process should get stuck there. Check how many entries do you have in VBDATA currently. You might want to run the reorganize updates program to clear the old updates that are pending/in error (as these are from production).

Other than that, it might be to with deadlocks. Check the usertrace folder for any files when this happens the next time. Do you have any errors in alert<SID>.log?

You can also execute SM12 - Diagnostics in Update to see if all is well with the system. System log also might have some clues.

Siva

[Former Member]

Hello Siva,

This is what I did:

> Check how many entries do you have in VBDATA currently.

More than 1.000 in every SAP systems that Iu2019m creating from Production

> You might want to run the reorganize updates program to clear the old updates

> that are pending/in error

Good point!!

After the reorganization (ST14 u2013 Reorganize update requests) the table VBDATA is empty.

Iu2019ll check and Iu2019ll let you know if the problem is definitively solved.

Thanks for you help,

Federico

[Former Member]

Guys,

I solved my problem with an Oracle patch: p5092134_10202_Linuxx86-64.zip

Patch 5099995 : applied on Wed Jan 14 10:25:02 PST 2009

Created on 16 Mar 2006, 13:29:18 hrs US/Pacific

Bugs fixed:

5092134

Thanks for your help and your time,

Federico Biavati