WebSphere Application Server

Page navigation

WebSphere Application Server Horizontal versus Vertical JVM Stacking Report

This report presents performance data for a test that compares the effects of stacking 200 WebSphere Application Servers.

The question is whether it is better to host all WebSphere Application Servers on a single (vertical stacking) z/VM guest, or distribute the servers across multiple guests (horizontal stacking). As an additional test, the use of shared mini-disks or a DCSS for the WebSphere installation was analyzed, to determine if there are criteria that favor one or the other setup.

Creation date
February 2012

WebSphere on IBM System z 64-bit/31-bit Studies with J2EE Workloads

This study explores the performance of a WebSphere Application Server 6.1 system under a customer-like J2EE application workload. It includes a very detailed description of how the test environment was set up and how the systems were configured. The difference in performance behavior of the 31-bit and the 64-bit WebSphere versions is compared, and the impact of heap size and garbage collection is analyzed.

It was found that with a special scenario at the highest workload level, which utilized eight CPUs, the 64-bit WebSphere version with a large heap showed its strength. The CPU scaling is very linear, but with the high computing power of the IBM System z10, the network bandwidth becomes a critical factor. At the highest workload submission rate, for example, a 10 Gb Ethernet card was needed to manage the traffic from the workload generating clients.

Creation date
September 2009

Tuning WebSphere Application Server Cluster with Caching

This paper analyzes parameter and configuration variations in a WebSphere Application Server cluster running the Trade workload when caching is enabled. We used a secure environment with a DMZ to protect the application servers against an uncontrolled external zone.

These are the most interesting findings:
  • Any form for caching resulted in a significant throughput improvement over the no caching case, where Distributed map caching generated the highest throughput improvement.
  • DynaChache disk-off load is a very interesting feature, which can significantly improve the performance with small caches without additional CPU cost.
  • The z/VM VSWITCH LAN configuration resulted in higher throughput than the Guest LAN feature.
  • The IBM System z10 system obtained a significant throughput advantage over the IBM System z9 system.
  • We can highly recommend z/VM for environments such as this.
Creation date
December 2008

WebSphere Application Server 6.1 Base Performance

The paper gathers Linux end-to-end performance measurements for a WebSphere Application Server 6.1 environment. It shows how the set of products that is needed to run the Trade 6 benchmark on Linux for System z9 performs on a release-to-release basis.

These are the most important findings:
  • RHEL 5.0 achieves a very good CPU scalability
    With 8 CPUs on the WebSphere LPAR we obtained a scaling factor of 7.5 times.
  • We recommend using RHEL 4.5 or higher for WebSphere 6.1.
Creation date
September 2008

WebSphere Application Server Base Performance

The paper gathers Linux end-to-end measurements for all of the components in the path from the user accessing the WebSphere Application Server system to the database.

The paper shows how this set of products performs on a release-to-release basis, and how the performance can be improved.

The components studied are:
  • SUSE Linux Enterprise Server (SLES)
  • WebSphere Application Server
  • Java
  • IBM DB2 Universal Database on Linux for IBM z Systems or on z/OS
Creation date
December 2007

End-to-End Performance of a WebSphere Environment Including Edge Components

This paper describes the end-to-end performance of a WebSphere Application Server 6.0.2 cluster environment, including firewall systems, Edge components such as WebSphere Load Balancer and the caching proxy server, and the Web server.

It uses a typical setup for application servers providing services into the Internet. The systems with the application server are protected by implementing a demilitarized zpnme (DMZ), which suppresses direct access from the Internet completely and allows only well controlled accesses to the servers via the proxy. The entire environment was set up with z/VM guests on z/VM 5.2.

Through the project, we gathered experience on how to analyze and tune a cluster environment on Linux on IBM z Systems using the Trade 6.0 (Trade) application. We also learned where some of the pitfalls can be found. We were trying to find the best configuration to achieve the highest throughput.

Creation date
January 2007

How to Determine the Causes of Performance Problems with WebSphere Applications running on Linux for zSeries

What do you do when you get a call from your customer, who says this:

“We have a WebSphere application running on Linux for zSeries, and its performance is bad”.

The following document describes a process to follow to help diagnose performance problems, and suggestions for resolving them.

The document is intended to be used by IBM field personnel working with the customer to diagnose such a problem, or by the customers themselves.

Creation date
July 2004