EHP6 Upgrade - Checks after phase MAIN_TRANSEXEC/TABIM_TRANS were negative!

[Former Member]

Hi,

I have got an error in the execution phase MAIN_TRANSEXEC/TABIM_TRANS-

4 ETW000 date&time: 08.03.2013 - 22:37:11

4 ETW000 2 errors occured.

4 ETW000 end of log fragment of child 00000029684
4 ETW000 child process rc = 12
4 ETW690 COMMIT "23193241" "274866271"
2EETW000 child process 31784 terminated -> setting rc=12 for this import portion

4 ETW000  [     dev trc,00000]  Transaction (Process ID 70) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.

4 ETW000                                                                                                  53  112.367544

4 ETW000  [    dbds    ,00000]  ***LOG BY2=>sql error 1205       performing EXE                           41  112.367585

4 ETW000  [    dbds    ,00000]  ***LOG BY0=>Transaction (Process ID 70) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.

4 ETW000                                                                                                  47  112.367632

1AETW000 SQL error 1205 occured: Transaction (Process ID 70) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.

2EETW000 Please contact the SAP support.

Windows 2008 R2

MSSQL 2008 R2

Regards,

[Former Member · Accepted solution]

Solved : Set the deadlock priority to Low.

Solution – Set the Dead lock priority to low

 DECLARE @deadlock_var NCHAR(3);

SET @deadlock_var = N'LOW';

SET DEADLOCK_PRIORITY @deadlock_var;

GO

After the upgrade –

Set the deadlock priority to NORMAL.

SET DEADLOCK_PRIORITY NORMAL;

GO

[Former Member · Accepted solution]

Hi,

Good that you have solved the problem... btw could you please tell me what have you referred in order to solve it ....

Thanks and regards

Raghu

[Former Member · Accepted solution]

Thanks Letz.  I'm running into the exact same problem with EHP6 upgrade.  I'm trying out your solution.  I will report back. 

[Former Member · Accepted solution]

No luck with your fix.  Did you have to do something special to set that deadlock priority?  I set the deadlock priority to low and selected "repeat phase  MAIN_TRANSEXEC/TABIM_TRANS to continue at the point it stopped".  Should I have selected to initialize from the beginning of the phase instead?

[Former Member · Accepted solution]

Execute it in MSSQL Management Studio

[Former Member · Accepted solution]

Hi,

Update your R3trans as per the below note and see whether it gets solved.

Note 1127194 - R3trans import with parallel processes

On MS SQL Server and with the 720 Kernel, R3trans version of 05.02.13 or later (kernel 720 patch number >= 420) is recommended in order to further accelerate the import by avoiding deadlocks on table REPOSRC caused by an optimizer bug (instead of automatically handling those deadlocks).

Regards

Raghu

[Former Member · Accepted solution]

Hi,

I also faced the same problem today. And it got solved.

Solution applied.

1. Stop SUM ...** Caution: Dont Reset

2. Stop SAP Services

2. Download latest files of R3trans (421), tp(420) and lib_dbsl(420) and replace them in the kernel folder.

3. run SAPCPE for sync.

4. start SAP Services. ** Caution dont start SAP

5. Start SUM and repeat the Phase.

It is working fine..

Regards

Raghu

[Former Member · Accepted solution]

Thanks for your help.  I was able to go back and solve the problem by redoing the upgrade (this was a demo/test system) and instead of selecting "single system" I did the upgrade as "downtime minimized".  This problem seems to happen with the "single" system option more often than not. 

I also found the following on the SUM release notes for SQL server (OSS 1765220): "SQL Server deadlocks may occur. The abap runtime errors will be DBIF_REPO_SQL_ERROR indicating SQL error 1205 when accessing program RSDBRUNT. These errors may be ignored."

So I guess I could have also hit ignore and continue on. 

[former_member413692]

Dear all,

For your information, I have tried setting the deadlock priority, downgrading the tp version, etc. But deadlock will just keep going.

We have used another server for further experiments and our conclusion is: what you need to do is, simply click "Continue" whenever you meet a deadlock. Deadlocks are currently unavoidable.

[Former Member]

Same issue here. so to fix it I didnt changes the deadlock priority. I just continue whenever I got the deadlock.

Thanks,

Rajesh