Extraction process is taking long time

[Former Member]

Hi all,

We are using 0fi_gl_14 DS in our process chain and sceduled it in every 2 hrs, Previously it is working fine. But now it is taking too long time to run extraction process (job) in r/3 side and for extracting 3 records it is taking 2 to 3 hrs and it is overlapping with the next schedule.

Could you please let me know what might be the reason, and how to improve the performance?.

Thanks,

Harish

Edited by: harishhh on Jul 12, 2011 8:59 AM

[Former Member · Accepted solution]

Hi Harish,

Are there any Index created for this data source in ECC.Creating INdex will help you in this situation .Take the base tables for this data sources and create Index on those Tables .

Regards,

Amit Trivedi

[Former Member · Accepted solution]

Thank you for your suggestion Amit ,

The datasource 0FI_GL_14 extracts data from the tables FAGLFLEXA BKPF, BSEG, and BSEG_ADD, all these tables has index ,in which FAGLFLEXA has index creation according to note 1135916 .

could you please suggest me any other suggestions?

Thanks,

Harish

[former_member186445 · Accepted solution]

do you have a user exit?

is this the only ds with bad performance or are different datasources impacted? maybe you don't noticed yet because they don't run every 2 hours like the one posing a problem

M.

[Former Member]

Hi Harsha,

Check in R/3 process are available or not. also any IDCOS is taking the time or any customer exists are there. if your not finding any this of this. Create Traces and check once again.

Thanks and Regards,

Venkat.

[Former Member]

Examine the SQL that's being executed on the R3 database. We had a similar issue, and then created a custom index on the R3 database. It was a concatenated index, with one of the primary fields being a timestamp used by the extractor for purposes of identifying the correct delta period, yet had no index in R3. The exact fields (and field order) may be somewhat dependent on your specific GL setup.

Edited by: DSchiffhauer on Jul 18, 2011 1:57 AM

[Former Member]

Hi All,

Thanks for your valuable suggetions.

There is a BADI devoleped to the data source (0FI_GL_14).

How to analyse the performance of the BADI ??

This weekend we are going to rebuild the indexes for all tables ralated to the data source 0FI_GL_14, This may improve performance.

I welcome any other suggetions...

[joaquim_casasbaca-castex]

Hi harish, at the end the performance improved ? and if so, do you remember which were the new indexes created and in which tables ? I appreciate your comments and suggestions.

Thanks in advance.

Regards

[Former Member]

2 hours for 3 records... something horribly wrong.

Can u please check if enough background work process are available in ECC.

Do one thing..go to monitor. > Environment-> job overview>In the source system.

check how much time it took and try to analyse it.

also take job name and go to sm37 and see id there was any delay for this job. if these is delay then it means BGD process were not available.

From job analysis you can make out where it took time.

check this and give us more details. We will be able to help you.

Regards

amandeep

[Former Member]

Hi Aman,

Thanks for your suggestions.

When I analysed the Job I found out that the delay has occured in executing Function module RSC2_QOUT_CONFIRM_DATA which took 3 hours for its completion .

Could You be please analyse what has caused this delay ?

Thanks alot Aman

[Former Member]

Otherthan these notes, Please share me the info..

The list of performance tuning activities to reduce the extraction process (job) running time?

Edited by: harishhh on Jul 13, 2011 9:05 AM

[Former Member]

Hi,

Also check the solution given in the note: 1531175.

Also check the note :1322418

Regards,

Anil Kumar Sharma .P

Edited by: Anil Kumar Sharma on Jul 12, 2011 1:16 PM

[Former Member]

We have done SQL trace using st01 and analysed it..

long running job for access to table FAGLFLEXA is taking more time in database.

Please suggest us with further solution.

Regards,

Hari

[Former Member]

Here is the job log I highlighted where it is causing problem

01:12:25 Job started

01:12:25 Step 001 started (program SBIE0001, variant &0000000029033, user ID ALEREMOTE)

01:12:25 Asynchronous transmission of info IDoc 2 in task 0001 (0 parallel tasks)

01:12:25 DATASOURCE = 0FI_GL_14

01:12:25 RLOGSYS = ******

01:12:25 REQUNR = REQU_9A***

01:12:25 UPDMODE = D

01:12:25 LANGUAGES = *

01:12:25 *************************************************************************

01:12:25 * Current Values for Selected Profile Parameters *

01:12:25 *************************************************************************

01:12:25 * abap/heap_area_nondia......... 2000683008 *

01:12:25 * abap/heap_area_total.......... 2000683008 *

01:12:25 * abap/heaplimit................ 40894464 *

01:12:25 * zcsa/installed_languages...... ED *

01:12:25 * zcsa/system_language.......... E *

01:12:25 * ztta/max_memreq_MB............ 2047 *

01:12:25 * ztta/roll_area................ 6500352 *

01:12:25 * ztta/roll_extension........... 4000317440 *

01:12:25 *************************************************************************

01:12:26 1 LUWs confirmed and 1 LUWs to be deleted with function module RSC2_QOUT_CONFIRM_DATA

06:40:50 Call customer enhancement BW_BTE_CALL_BW204010_E (BTE) with 6.006 records

06:40:50 Result of customer enhancement: 6.006 records

Can someone suggest me what needs to be done to get rid of this problem..

[Former Member]

pls check the following notes:

1416253 & 1432562

rgs,

Parth.