Skip to Content

Updated: the Java project has been uploaded in sdn.sap.com.

Related weblogs: Simple ABAP Benchmark and Simple SAP Benchmark – ABAP vs. Java

Here is the Java version of the weblog Simple ABAP Benchmark

As an SAP Technical consultant I’m often involved in tuning projects.

After logging into an SAP system I’m curious to know if the system is basically a powerful one or not.

It is often useful to compare DEV with QAS with PRD systems and also the different application servers of the PRD system.

I decided to share a Java Web Dynpro project.

I try to measure:

  • the CPU looking for primes with a very simple logic
  • the disk speed writing 100MB 10 times in the work directory
  • the RDBMS writing a small record into a table
  • the RAM allocating a huge internal table and than scanning it sequentially

I do not try to measure:

  • CPU numbers
  • Application numbers
  • Buffer quality

The Web Dynpro Java projects (Dictionary / Web Dynpro) can be downloaded here http://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/com.sap.km.cm.docs/library/uuid/10e2c653-f32a-2a10-77ac-c4722746b6d8

Few results follow:

Linux

  • CPU highest prime: 20.935
  • File system – MB : 2.060
  • RDBMS – records : 16.900
  • RAM – scans : 13.193

Laptop

  • CPU highest prime: 24.689
  • File system – MB : 370
  • RDBMS – records : 4.300
  • RAM – scans : 5.178
To report this post you need to login first.

7 Comments

You must be Logged on to comment or reply to a post.

  1. Hi Sergio,

    We recently found out that the performance of CPU’s is also measured in terms of SAPS by SAP. In my recent project we had to tell SAP the power of our CPU’s in terms of SAPS which is “The SAP Application Performance Standard (SAPS) is a hardware-independent unit that describes the performance of a system configuration in the SAP environment. It is derived from the Sales and Distribution (SD) Benchmark, where 100 SAPS is defined as 2,000 fully business processed order line items per hour.”

    Have you ever come across this? Also can you also provide some program for us to measure this.

    Your weblogs have been a great help to us.

    Thanks for the same

    Regards
    Sidharth

    (0) 
    1. Sergio Ferrari Post author
      Hi Sidharth,
      I’m happy that the weblogs were interesting.

      About SAPS:
        it is exactly what I was looking for…
        As you know it’s a great way to measure a SAP system but it requires a effort to set it up and running.
        In add it was perfect at times of SAP R/3 but what about CRM, APO, SRM, BI, standalone ABAP Application Servers?
        I think a few others SAP Benchmark are available (at least for APO) but I never heard anything about SAPS related to SAP J2EE and so what about our SAP NetWeaver Portals?
        About J2EE there are a lot of non-SAP specific benchmarks that should work fine (for example see http://www.spec.org/benchmarks.html) but again how much effort is required and how to compare them with ABAP systems?

      “So, this weblog series about Simple SAP Benchmark is really helpful”

      (0) 
      1. Birgit Sabaschus
        Hi Sergio,

        See also my comment in your “Simple ABAP Benchmark” weblog. SAP offers a variety of SAP Standard Application Benchmarks to cover different kinds of business scenarios. You can visit SAP’s public web site SAP Quick Sizer 2005 can calculate SAPS recommendations for sizing the NetWeaver Portal.

        Kind regards,
        Birgit Sabaschus

        (0) 
        1. Birgit Sabaschus
          ooops, something happened to my posting
          next try:

          Hi Sergio,

          See my comment in your “Simple ABAP Benchmark” weblog. SAP offers a variety of SAP Standard Application Benchmarks to cover different kinds of business scenarios. You can visit SAP’s public web site http://www.sap.com/benchmark to obtain information about the available SAP benchmarks (for example, BI, APO, TRBK, Portal…).
          SAP provides two benchmark scenarios running on the NetWeaver Portal. Unfortunately no Technology Parter has certified a portal benchmark so far.
          You do not have to run the benchmarks by yourself. You can use the certified benchmark results at http://www.sap.com/benchmark. If your hardware is not in the list you can ask your hardware vendor for the corresponding SAPS value. Most of the partners run internal load tests to get the SAPS values of all of their servers.
          Business processes of different solutions (CRM, SCM, SRM,…) finally all meet on the persistence layer of the SAP NetWeaver Application Server causing standard operations such as INSERT, UPDATE, DELETE. There are two flavors: ABAP and Java. If you run benchmarks and performance measurements on the same hardware you can receive load factors and algorithms for the different solutions and you can relate this to SAPS. For example the Quick Sizer – Getting started can calculate SAPS recommendations for sizing the NetWeaver Portal.

          Kind regards,
          Birgit Sabaschus

          (0) 
  2. Thomas Szücs

    Hi,” instead of “into” to save copying the whole line Moreover, as you used int for the key in Java you could so as well in ABAP and replace the n-type by i. As you seem to be more interested in accessing a single field using “read table .. transporting ..” could be a solution as well.

    (0) 
    1. Sergio Ferrari Post author
      Thomas, thanks for the suggestions…
      In Java, I’m going to implement the list instead of the vector.
      In ABAP, even if you are right, I do not intend to make it faster.
      The goal of the test is to stress the system also forcing more work then necessary in the specific case but usually required by business process…
      (0) 
    2. Sergio Ferrari Post author

      Thomas, I decided to accept your suggestions also about ABAP, so I tried both options:Changing the type of the key field improved the performance of 5%.

      (0) 

Leave a Reply