topic Re: ABAP performance in Application Development and Automation Discussions

ABAP performance

Former Member — Wed, 09 Apr 2008 07:13:15 GMT

Hi!

I have an internal table, which has 250000 records.

LOOP AT itab INTO wa.
...
* here are some other ABAP statements, but no SELECT
* or other database accesses
...
READ TABLE ...
MOVE ...
...
ENDLOOP.

This code part runs for 8 minutes and I don't know why is it so slow.

How can I trace, which command takes the most time? If I go into the debugger and run it step-by-step, it seems every command is progressed instantly. Is there some tool for it?

SE30 is not enough detailfully, so it is not useful in my case.

Thank you

Tamá

Re: ABAP performance

Sm1tje — Wed, 09 Apr 2008 07:15:28 GMT

Maybe the read table is the bottle neck. How did you declare the internal table you are reading here?

Standard, sorted, hashed table, with (non) unique key?

Re: ABAP performance

ThomasZloch — Wed, 09 Apr 2008 07:20:10 GMT

yes, you have to tell us more.

Most likely your READ statement does not use a qualified key. Also, you might want to drop "INTO wa" and use field symbols instead ("ASSIGNING <fs>"). At 250K records you're saving some copying costs.

Cheers

Thomas

Edit:

You have to explicitely switch on analysis of internal table operations in SE30. Also see this excellent blog:

/people/siegfried.boes/blog/2007/11/13/the-abap-runtime-trace-se30--quick-and-easy

Re: ABAP performance

Former Member — Wed, 09 Apr 2008 07:29:28 GMT

Hi,

u can use ST05 <Tcode> for detail check

Regards,

sathish

Re: ABAP performance

Former Member — Wed, 09 Apr 2008 08:17:13 GMT

I can't access that system now, I'll popup later.