Performance of report

[Former Member]

Hi All,

Execpt programming code is there any way to improve the performance of the Report.

Thanks,

Subbu.

[Former Member]

Hi,

Yes, make sure that your database is in good shape by regularly archive data and reorganize your database tables.

Naturally it helps if the hardware is sized to deal with the workload, both database and application servers. But this is usually a lot more expensive than creating performant code.

Regards,

Stefan Ericsson

[Former Member]

Hi,

you can check the performance of your code using [SE30|https://www.sdn.sap.com/irj/sdn/weblogs?blog=/pub/wlg/7769] [original link is broken] [original link is broken] [original link is broken]; and [ST05|https://www.sdn.sap.com/irj/sdn/weblogs?blog=/pub/wlg/7205] [original link is broken] [original link is broken] [original link is broken];. the first one is runtime analysis , indicates which part of you code is taking time, abap or database or application. then st05 will indicate the problematic and time consuming SELECT statements.

additionally chk this too: [abap tips|https://www.sdn.sap.com/irj/sdn/wiki?path=/display/abap/abap%2bperformance%2band%2btuning]

regards,

madhumitha

[Former Member]

To improve the Performance in a report..

1.Dont use corresponding in the select statement.

2.Dont use select statement inside the loop.

3.Use binary search while using read statement.

4.In the select statement use the proper key fields.

this will help u......

Plz Reward

Regards

Anbu

Edited by: anbu b on May 22, 2008 12:22 PM

[former_member251078]

Hello,

ST05 and SE30 are became old now!!!.. you can use ST12 transaction for taking both abap and sql trace. Normally if you are finding most of the time is being spent in database, they you have to analyze the expensive sql statement in sql trace of ST12 transaction. There are various options in ST12 trace tool. Please have a look into SAP Note 755977 for details.

Yours Sincerely

Dileep

[former_member194613]

> ST05 and SE30 are became old now!!!..

the last posting is absolutely misleading, the two traces are still up-to-date and the ST12 is just something on-top of the traces.

For a basic check a simple usage of the traces, as described here is absolutely enough:

SQL trace:

/people/siegfried.boes/blog/2007/09/05/the-sql-trace-st05-150-quick-and-easy

SE30

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

Siegfried

[Former Member]

Hi Subbu,

as Stefan already mentioned, you can improve the performance by bigger or faster hardware. One of my professors always told us: "You could always improve your software my a factor of 1000. You can raise the clock rate by 10, you can achieve a factor of 10 by changing the architecture and you can improve you algorithm by 10."

I think this is still true for most software packages. The factor of 10 in all areas is probably not always possible, but it shows in which directions you could go.

The bigger of faster hardware is equivalent to the higher clock rate (In SAP environment memory could play a comparable role than clock rate). The algorithm is equivalant to code changes (you ask for neglecting it).

The architectural change you could do is to go from two tier to tree tier or vice versa. If you have a report which accesses the database with lots of small requests, you may see a big improvement if you run the application on the database server. May be you find other reasons to make architectural changes in your environment. Not all of them needs to be more expensive than your current setup.

May be this could be some small hint to find a solution for your problem, even you haven't told us what it is.

Best regards

Ralph

[Former Member]

Hey Ralph,

Thanks for your Idea. But it is not that much easy to change the architecture of servers...

Thanks,

[former_member194613]

> Execpt programming code is there any way to improve the performance of the Report.

the performance of your report can not be changed by changing the hardware, because strictly speaking I would say the performance of a report is a property of the report and If you measure the time you would have to specify the power of the hardwarfe also (the SAPS on the benchmark pages is a good measure).

There are of course several possibilities the change the performance of a report without touching the code:

+ check whether all used and necessary indices are optimal, if not optimize them

+ check the buffer settings of the used tables, maybe some tables could alos be buffered

+ check the buffer size, is it swapping

There are probably some more. But I think it is enough for the game, because I would never recommend to see it separately. You should consider coding changes and system setting together.

Siegfried

, but it is

[Former Member]

Thanks Siegfried,

But i heard like we have Parllal Processing and cursor methods to improve the performance of the report.

Thanks,

Subbu

[former_member194613]

and they work without coding changes ..... no!

As I have said, you should see the whole picture, not coding and system separated.

Parallel Processing: Optimizes response time, but does not optimize resource consumption, because parallel processing has extra costs.

Cursor ... what do really mean ... rarely necessary.

[Former Member]

Check Tcode : ST05 which tells u the load on the database .

SE30 is for program performance anaylsis