Incorrect number of data lines in TST03

[Former Member]

Hi,

one of our scheduled jobs is being cancelled with no errors but those found on SP12 - Consistency Check:

Incorrect number of data lines in TST03

Any idea of what might be happening?

SM21 reports the following:

Error 13 for write/read access to a file. File =

Error at DB commit, return code 016384

SAP Commit Failed

Stop Workproc20, PID 9915

Start Workproc20, 4 times since system startup. PID 10255

Invalid object name for TemSe object: " "

Failed to create log for job /SAPAPO/BACKGROUND_SCHEDULING

Calling program reports invalid handle for TemSe object (magic==X'NULL-ptr')

Delete session 001 after error 023

Thanks

Antonio

[Former Member · Accepted solution]

Hi Antonio,

Basically TST03 is a table related to TemSe data.

I hope may be while the background job is trying to access the table for read/write operations there is some commit work got failed.

Check whether that table TST03 is locked.

If this is locked .unlock it and perform the operation once again.

Thanks & Regards,

Sudheer.

[JPReyes]

Have you tried to run a TEMSE consistency check and removing inconsistencies before running the job again?

Regards

Juan

[Former Member]

Hi, thank you both for your answers...

The table TST03 is not locked and I've eliminated the inconsistencies reported by SP12's Consistency Check, however the error remains and a new inconsistency is added upon every job execution...

[Former Member]

Hi António Barrote

I beleive that you found the solution for this issue, could you please share the solution since i too face the same issue

regards

Naveen

[Former Member]

Hi Naveen...

yes, the problem was solved but it had nothing to do with SAP itself...

We have our APO system on two servers: the message server (and DB) runs on an iSeries and the application server and LiveCache server on a SuSE Linux machine. This last server has several iSeries shared (NFS) folders mounted and all TemSE objects are created on one of these folders.

Now the problem is we use a Vision solution for our HA. Whenever Vision is "touching" a file, Linux can't "see" the lock status due to NFS limitations. So, when it is trying to access the same file, a "permission denied" error message is issued to SAP, causing the job to abort and the TemSE object to become inconsistent.

One solution is to change from NFS to Samba.

Another one, much faster to implement, is stopping the Vision job during the night period, when those SAP jobs run. Since then, we didn't had any more problems.

Of course, this is simply a workaround and not a permanent solution!

Could this be your scenario as well?

I hope you can solve your problem...

Merry Christmas and a Happy 2009 for everyone!

Antonio

[Former Member]

Hi António Barrote

Thanks for your detailed explanation, my scenario is different.

I work in Solution Manager on Solaris server, here all the SAP background jobs are getting canceled without any logs. After exploring, i find that TemSe objects become inconsistent.It shows like 70000 objects are inconsistent.

In some of the objects it has the comments as Incorrect number of data lines in TST03. Now i dont know how to proceed.

Also i dont know whether reorganization of the table TST03 using BRTOOLS will help ?

If you have any suggestions, please post them

regards

Naveen

[Former Member]

Hi Naveen,

70.000 inconsistent objects is too much, you're right...

Run report RSTS0020 on the background to find out the list of all inconsistent objects.

Then, run report RSTS0030 to delete all those inconsistent objects (select TST01 and TST02).

Do you have all your SAP standard reorganization jobs running?

Good luck

Antonio