SAP SCM LiveCache SQL error "-60" occurred while executing Native SQL after migrating to EHP3

[former_member315367]

Hello,

We face this error on our SCM system after migrating to EHP3 (from EHP2). Short DUMP details :

Error "-60" has occurred on the current database connection "LCA".

Database error text: POS(1) Work rolled back: ERROR 30709 in LCK_ENQUEUE02: S

Triggering SQL statement: "EXECUTE PROCEDURE LCK_ENQUEUE02"

LiveCache version is : 7.9.08.29 / Kernel    7.9.08   Build 029-123-252-402

SCM version is 713 SP8

Does someone faced this error before?

[Former Member · Accepted solution]

Hello,

Thanks all of you for your feeback.

It seems that the issue is link to the parameter LCA_LC_LCK  = LCA in user parameter.

When we remove from user parameter LCA_LC_LCK, we don't have such issue and we can acces to planning book.

Do you know what's the aim of this parameter?

Best Regards

[former_member229109 · Accepted solution]

Hello,

- I assume, that you reported the case in SAP messages 383225 / 2015 and 493688 2015.

  Correct?

- In case of further questions, you could update the message 493688 2015.

- As it was written in the SAP SCM documentation, that there are 3 connections created on the SCM system - LCA, LDA and LEA. Those connections are pointed to the same liveCache instance & used by applications to connect to the liveCache.

when you run a SCM system copy, those connections need to be checked, changed, if it will be needed on target system, see SAP note 886103.

LEA connection is used by lock server applications. As of Release SCM 5.1, Demand Planning uses the lock server based on liveCache. (Up to Release SCM 5.0, it uses the Web Application Server (Web AS) ABAP enqueue server.) The lock server based on liveCache can manage multidimensional locks.

Additionally review the SAP notes: 1950649 - LiveCache lock handler not available

** The user parameters started with "LCA_LC" should not be set in a customer's environment. By parameter LCA_LC_LCK you set the liveCache connection to lick server to LCA connection, and it is NOT recommended.

Best regards, Natalia Khlopina

[babu_kilari4]

Hello Benoit,

We did face similar error (but not on Enqueue level...but the error code is still the same SQL -60) after upgrading to SCM 7.0 EHP3 recently and SAP had recommended to run the program /SAPAPO/OM_CHK_REBUILD_TQA_ADM on a daily basis after business hours. Also, they did state the error can also occur for various other reasons. As mentioned by one of the experts, you need to raise the issue to SAP component area BC-DB-LCA via Incident ticket.

Thanks & Best Regards,

Babu Kilari

[former_member229109]

Hello,

It's the general MAXDB error:

What does SQL error -60 mean?

              SQL error -60: Work rolled back 

              The transaction was implicitly canceled and reset due to an implicit rollback.

=>

Pay attention to the section  "Database error text"  in the ABAP dump for the analysis of the issue with the failed liveCache transaction.

In this case:

Database error text: POS(1) Work rolled back: ERROR 30709 in LCK_ENQUEUE02: S

Triggering SQL statement: "EXECUTE PROCEDURE LCK_ENQUEUE02"

So check the LCA RC 30709 in //om10, as I wrote above, The LCAPPS routine  LCK_ENQUEUE02 failed with this RC.

Regards, Natalia Khlopina

[former_member229109]

Hello,

I recommend to create the SAP message to BC-DB-LCA component for the analysis of "ERROR 30709 in LCK_ENQUEUE02".

Error 30709 in LCAApps routine, check hint of the RC in /n/sapapo/om10, Attempt to Narrow Time Horizon(ERR_CONTROL_LOW_TIME_LIMIT), the Year 2038 Limit.

It will be helpful to check the LCA traces & get details on the appl. scenario which was failed with the reported ABAP dump.

Regards, Natalia Khlopina

[JamesZ]

hi，

-60 indicates deadlock, do you know how to reproduce this issue? Also can you attach the who dump to this discussion?

Best regards,
James

[former_member315367]

Hello,

Yes I can reproduced this issue every time in /SAPAPO/SDP94 on all migrated systems (EHP3). Dump file is now attached. Kepp me posted.

dump.txt.zip

[JamesZ]

Hi,

In that case, please also provide maxdb logs as below:

1. log on maxdb host and execute below command:

dbmcli -d <dbsid> -u <username>,<password> diag_pack

Afterwards you will find a file name diagpack.tgz under maxdb run directory, so please attach it to this message.

Best regards,
James

[former_member315367]

Hello James,

Please see the diag pack attached (please remove txt extension).

diagpkg.zip.txt.zip

[JamesZ]

Hi ,

Unfortunately the maxdb kernel log does not record this deadlock. Please attach the whole short dump. Also attach the corresponding sap work process trace for the dump.

Cheers,
James

[former_member315367]

Hello,

Please see attached files. I'm not sure about "whole" short dump, do you need more information?

dev_w7.txt.zip

dump.txt.zip

[JamesZ]

Hi,

Please check if the problem gets solved via turning off locking of liveCache:

1. t-code /sapapo/msdp_admin
2. deactivate the liveCache based locking for each planning area separately in the section "Locking Logic".

(neither by batch jobs,

nor by interactive users, nor by CIF transfers).

Please note:

!!!!!!!!!!!!!!!!!!!!!!!!!!!!

When the locking logic is being turned off, there must be no access to the corresponding planning
area (neither by batch jobs, nor by interactive users, nor by CIF transfers).

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

You may need to turn locking off for all planning areas, then restart liveCache.

If above works, then the permanent solution or the prerequisite to finding permanent solution is to update latest LCA build according to SAP note I would like suggest you updating the latest liveCache build as per SAP note 2127659.

Best regards,
James

[former_member315367]

Hello James,

We have untick the "activate liveCache Lock" and now we can modify the planning book. What's the aim of this flag? This flag is there since the go live in 2011 and we never have an issue except now during the migration from EHP2 to EHP3 in SCM.

By the way we are already on last LiveCache version (package LC790829A1024_3-10009612.SAR).

Thanks for your help.

Regards,

[JamesZ]

Hi,

The aim of the lag is for whether you can activate liveCache lock. Turning it off means you are going to use scm server lock. The liveCache lock
is smaller than scm server lock. If you are already using the latest LCA build, then this issue may be a bug.

I would like to recommend you raise a ticket to sap component bc-db-lca for a permanent solution. However in the meanwhile, you can keep my workaround as a temp one.

best regards,
James