DevOps and System Administration Forum
cancel
Showing results for 
Search instead for 
Did you mean: 
Read only

SAP Performance Benchmark

lakatbalazs
Explorer
0 Likes
3,034

Hello!


I am looking for a way to benchmark performance between our SAP Systems.

The ST03 data is not suitable for this, because it shows averages that are dependent on the usage profile (poorly optimized Z reports are making the avg. resp. time look bad )


So is there a way to monitor the SAP System performance independent of the usage?


Regards,

Balázs

10 REPLIES 10
Read only

S_Sriram
Active Contributor
0 Likes
2,641

Hi Balázs

For the Z- reports you can enable the trace using the transaction code ST12 and do the fine tunning and performance of the reports

Regards

SS

Read only

0 Likes
2,641

Hi,

thanks, but optimizing the Z report is not my problem here. My task would be to compare the system performance independent from the load. So basically how can I compare System A and B's performance independent from the transactions running on them?


Regards,

Balázs

Read only

S_Sriram
Active Contributor
0 Likes
2,641

Hi Balázs

If you want to get back the performance of your system, as a basis guy you have to analyze all the very bad performing business transactions and Z-reports, for this, you have to take the trace of the Z-report, check which SQL statement taking long durations to fetch the data from the table, 90% of the problem because of the index is not created the selection of the Sql statements.

For any ABAPer, while creating the Z-reports, they may forget to check the overall performance of the reports in the prod system at the same time it may be working fine in Dev & Qas systems, all the issues will appear in the production environment, so you have to create the trace and check with ABAPer to find the root cause.

Regards

SS

Read only

0 Likes
2,641

Hello,

thank you, but optimizing the performance is not my goal here. I want to compare two systems performance independently from the usage profile. So for example mesuere the SAPS value on them. Or run some kind of benchmark transaction or report which then gives a score to the system. As we do it for the graphics card for example 🙂

I was always presented with ST03 dialog response time charts, but they are representing the performance of the report and transactions not really the systems.

Regards

Balázs

Read only

amontella96
Active Contributor
0 Likes
2,641

lakatbalazs

I dont think there is such a System performance tool embedded in SAP ... do you have Sol Man? ...but still, I think the answer for you is lock the users (except yours 😉 ), stop the jobs that can interferes and test the response times running basic transactions/reports (you can also easily analyze your steps with SE30) hope this helps, cheers!A

Read only

0 Likes
2,641

Hello,

thanks but for a production system this is hard to menage. What basic transaction you mean?

I mean there has to be a standard way to do this. They also run some load when they test the SAPS value of a system.

We do have SolMan and it can run scripts on the systems but what to run exactrly?

Regards,

Balázs

Read only

former_member661012
Participant
0 Likes
2,641

Hello Balazs

From Standard way perspective, yes there are methods you can opt to optimize the performance.

If your problem is only specific to "Z Reports" then the tracing of the code becomes very important. For example, it is important for us to identify where the code is not written in an optimized manner and it is resulting in a bad response time.

To find out those things, you can use following transactions:

* ST12

* SQLM

* STAD - which should be compared with ST12(tabrec)

* /SDF/MON

* ABAPMeter

& if your database is HANA, then do execute the Mini Checks as well.

Reference Notes :

2436955 - How to collect and analyze traces using ST12 (single transaction analysis)

2524975 - Performance tracing overview: SAT, ST05 & ST12 traces

2953258 - How to chart ABAPMETER output when scheduled as a background job

https://archive.sap.com/documents/docs/DOC-47444

https://wiki.scn.sap.com/wiki/display/SAPSQL/How+to+run+transaction+st05+to+trace+a+program%2C+trans...

Hope this well help.

Regards

Basu Sharma

Read only

0 Likes
2,641

Hello,

thank you, but optimizing the performance is not my goal here. I want to compare two systems performance independently from the usage profile. So for example mesuere the SAPS value on them. Or run some kind of benchmark transaction or report which then gives a score to the system. As we do it for the graphics card for example 🙂

I was always presented with ST03 dialog response time charts, but they are representing the performance of the report and transactions not really the systems.

Regards

Balázs

Read only

S_Sriram
Active Contributor
0 Likes
2,641

Hi Balázs

In this case, you can execute the business transaction or report on both systems, using the transaction code STAD you can check the execution performance of the transaction in both systems and do the analysis or compare.

Regards

SS

Read only

0 Likes
2,641

Hello,

thank you, yes exactly but what transaction? SAP wrote here : "In technical terms, this throughput is achieved by processing 6,000 dialog steps (screen changes), 2,000 postings per hour in the SD Benchmark, or 2,400 SAP transactions.In the SD benchmark, fully business processed means the full business process of an order line item: creating the order, creating a delivery note for the order, displaying the order, changing the delivery, posting a goods issue, listing orders, and creating an invoice."
https://www.sap.com/about/benchmark/measuring.html

So there must be a standard way to do this, do you maybe know how?