topic Transaction SE30 usage in Application Development and Automation Discussions

Transaction SE30 usage

Former Member — Tue, 08 Aug 2006 13:00:52 GMT

Hi,

I am trying to analyse the CPU time taken for a "DO" statement using SE30.I want to compare this with the CPU time taken for a "WHILE" statement with the same logic.But when I execute for the "DO" statement multiple times I am getting different results each time which are highly varying.I am not sure which is the reliable or correct data.Hence could not further compare with the "WHILE" statement.

Cud someone suggest me how to compare performance with this?

Thanks in advance,

Hema.

Re: Transaction SE30 usage

Former Member — Tue, 08 Aug 2006 13:04:41 GMT

u can try this

v_time1 = sy-uzeit.

do.

enddo.

v_time2 = sy-uzeit.

dif = v_time2 - v_tim1.

Re: Transaction SE30 usage

Former Member — Tue, 08 Aug 2006 13:08:05 GMT

Hi hema,

1. I am getting different results each time

This is very much normal.

bcos the speed very much depends

upon the load on the server, at that particular point of time.

2. We can also measure in another way, programmatically.

GET TIME. "- for refreshing the time value.

T1 = SY-UZEIT.

DO.

*--- CODE

ENDDO.

GET TIME. "- again for refreshing the time value.

T2 = SY-UZEIT.

diff = T2 - T1. "GET DIFF IN SECONDS.

WRITE DIFF.

regards,

amit m.

Re: Transaction SE30 usage

Former Member — Tue, 08 Aug 2006 13:13:25 GMT

I think the best way to compare is trigger both the pgms at the same time and then check the cpu times.

Regards

Anurag

Re: Transaction SE30 usage

sridhar_k1 — Tue, 08 Aug 2006 13:33:32 GMT

Use statement GET RUN TIME in the program to get runtime of individual statement processing times instead of SE30.

Get run time also returns different times, take average

Try the following code.

data: rt1 type i.

get run time field rt1.

do 100 times.

....

enddo.

get run time field rt1.

write:/ rt1.