Maximizing performance

Many factors influence the performance of IBM® Rational® Quality Manager. With information about these factors, you can maximize the performance of your product.

About this task

Some of the factors that affect performance are as follows:
  • The browser that you use. For example, Rational Quality Manager works more efficiently on Firefox than on Microsoft® Internet Explorer.
  • The database that you use and how it is configured. The database software edition and version that you use can affect performance. For example, DB2® v9.5 performs better than v9.1. Also, DB2 Enterprise Server Edition performs better for larger installations than the Workgroup Server or Express® editions. Other database vendors might have similar software editions, for example, Enterprise vs. Express, which might affect performance in a similar way.
    Note: The Derby database that is included with Rational Quality Manager is intended for these situations:
    • Trial evaluations of the product
    • Deployment with test teams of 10 users or fewer with repositories smaller than 5 gigabytes
    Using Derby as a database for more users or larger repositories might affect performance.
  • The application server that you use. For example, using IBM WebSphere® Application Server as the application server can improve performance.
  • The number of test assets in the Rational Quality Manager repository. The use of server-side paging that is introduced in version 2.0.1 significantly improves performance for test teams with a large number of artifacts.
  • The server and users' network connection. For example, users or servers that are connected by LAN experience faster performance. Also, if the software system is installed as a dual-tier system and the servers are not installed in the same location, this might affect performance.
  • The software system topology. You can set up the software system as single- or double-tier. Separating the database activity from the Rational Quality Manager server positively impacts performance.
  • Server tuning could also affect performance.
  • Viewlet population. If you find that you do not use some of the viewlets on the home page, close them. The viewlets query the server frequently, which can reduce performance.

The following table contains links to detailed documents that describe how to maximize performance as you install and configure Rational Quality Manager:

Document Description Link
IBM Rational Quality Manager 2.0.0.1 Application Server Performance Report Examines performance and resource utilization of application servers Apache Tomcat versus WebSphere Application Server with the Rational Quality Manager Standard server in a single-tier topology. http://jazz.net/library/article/400
IBM Rational Quality Manager 2.0.0.1 Platform Performance Report Examines performance and resource use of Windows® versus Linux® platforms with theRational Quality Manager standard server in a single-tier topology. http://jazz.net/library/article/401
The Rational Team Concert™ (RTC) 2.0 Sizing Guide Provides guidance for enhancing the performance of Rational Team Concert. Because Rational Quality Manager is based on and integrated with Rational Team Concert, this sizing guidance might significantly impact the performance of Rational Quality Manager. This guide provides guidance for Rational Team Concert for hardware and software configurations that support these configurations:
  • Small scale enterprise configuration: Single-tier 100 - 700 users
  • Small scale enterprise configuration: Dual-tier 100 - 700 users
  • Large scale enterprise configuration: Single-tier 700 - 2000 users
  • Large scale enterprise configuration: Dual-tier 700 - 2000+ users
  • Multiple repositories (projects)

The document also provides artifact sizing guidelines.

http://jazz.net/library/article/205
IBM Rational Team Concert 2.0 Deployment Planning Guide Provides guidance for deploying Rational Team Concert so that it supports several thousand users on a single server instance. The document includes steps to carefully consider in these large deployments to ensure that the server capacity and tuning is appropriate for the anticipated user load. http://jazz.net/library/article/208

Feedback