SQL error 601 in SQL server 2012 SP1CU6

[former_member211576]

Hi experts,

   I just upgrade our SAP to Windows/SQL server 2012 and put the whole database on Fusion IO. We have a program running on SQL server 2008 without any errors. However, I found this program terminated with SQL error 601 every morning at 7:00am. It is weird that if we run it again later, it  complete without errors. We have check sap note: 965530 and it mentioned it happened only on older SQL server version. Please help.

---- ST22 --------------

Category               Installation Errors                                                         

Runtime Errors         DBIF_RSQL_SQL_ERROR                                                         

Except.                CX_SY_OPEN_SQL_DB                                                           

Date and Time          2013/10/08 07:00:08                                                         

Short text

SQL error 601 when accessing table "ZSCMPN".

What happened?

Error Text of the Database: "Could not continue scan with NOLOCK due to data

movement."

----  ABAP code --------------------

767	FORM GET_ZSCMPN .
768	"Exclude bulk and phantom P/N
769	SELECT Z~SITCOD Z~KUNNR Z~MATNR
770	INTO CORRESPONDING FIELDS OF TABLE I_ZSCMPN
771	FROM ZSCMPN AS Z INNER JOIN MARC AS C
772	ON Z~MATNR = C~MATNR
773	WHERE C~WERKS IN S_WERKS AND C~SOBSL <> '50' AND C~SCHGT NE 'X'
774	AND Z~SITCOD = P_SITE.
775
>>>>>	READ TABLE I_ZSCMPN INDEX 1.
777	IF SY-SUBRC <> 0.
778	"  IF I_ZSCMPN[] IS INITIAL.
779	WRITE:/ 'PAL P/N base is not exit! Stop the process!'.
780	STOP.
781	ELSE.
782	SORT I_ZSCMPN BY SITCOD KUNNR MATNR.
783	DELETE ADJACENT DUPLICATES FROM I_ZSCMPN COMPARING ALL FIELDS.
784	ENDIF.
785
786	IF S_MATNR[] IS NOT INITIAL.
787	LOOP AT I_ZSCMPN.
788	READ TABLE S_MATNR WITH KEY LOW = I_ZSCMPN-MATNR.
789	IF SY-SUBRC NE 0.
790	DELETE I_ZSCMPN.
791	ENDIF.
792	ENDLOOP.
793	ENDIF.
794	ENDFORM.

---  SM21 ------------

Details Page 2 Line 9 System Log: Analysis for all Remote Instances           1

Time	SAP instance	Type	Nr	Clt	User	TCode	Grp	N	Text
07:00:02	tccap45_TCP_45	BTC	028	218	76040301		BY	0	> Could not continue scan with NOLOCK due to data movement.

> Could not continue scan with NOLOCK due to data movement.                   

Details                                                                       

Recording at local and central time........................ 2013/10/08 07:00:02

Task......	Process	User......	Terminal	Session	TCode	Program	Cl	Problem cl	Package
03088	Background Processor No. 028	76040301		1		ZQP00759	X	Miscellaneous	SBAC

Documentation for system log message BY 0 :

Access to the database system has returned an SQL code indicating that

the operation could not be performed correctly by the database system.

This system log entry shows the database text for the SQL error that

has occurred.  It supplements the preceding system log entry, which

shows the SQL code, the table name and the SQL function where the error

has occurred.

Technical details

File	Offset	RecFm	System log	Grp	N	variable message data
259	409680			BY	0	Could not continue scan with NOLOCK due to data movement.

[sunny_pahuja2 · Accepted solution]

Hello,

> Could not continue scan with NOLOCK due to data movement.                   

It looks like that some other job/program is running at the same time. Could you please if some other job/program runs at same time which blocks access to table?

Thanks,

Sunny

[former_member211576 · Accepted solution]

Hi Sunny,

  The developer changes the code and hope it can eliminate this error. If the problem persists, I will ask him to reschedule the job.

[former_member211576]

Hi Sunny,

  Thanks,  Our solution is to reschedule the job at 7:05am and it works.

  By the way, is there any DMV can help me to identify which program is moving data while the program is scaning?

[clas_hortien]

Hi,

have you ran a DBCC CHECKDB as stated in the note 965530 ?

See note 142731 for details how to do it.

The error 601 was very often seen as a symptom of a database corruption.

Best regards

Clas

[former_member211576]

Hi Clas,

   My DB is pretty large(7TB compressed). It takes about 7 days to run a full checkdb in DOP=1. I checktable the table short dump indicated and it is ok. I select * from tcp.ZSCMPN(This table does not have secondary index) and it is no problem at all. Just as the note mentioned, if it is physical corruption, it will occur every time.

   Anyway, I am running checkdb with physical_only and it will take about 17 hours. Also, I takes a full backup of SQL 2008 DB now in case something goes wrong.

[Sriram2009]

Hi

Could you check the SAP 1250016 - Could not continue scan with NOLOCK due to data
movement

Thanks

Ram