Performance Tuning as an Art

There are as many ways to tune a Web server's performance as there are Web sites on the Internet. Depending upon the choices your company has made about its presence on the Web, you may be responsible for tuning your Web servers to best serve static Web pages or dynamically compiled application pages. Each type of site demands different hardware, application, and Windows 2000 and IIS performance tuning options. Another consideration is the amount of traffic that you may realistically expect your Web site to handle, particularly during peak load periods. Load will affect Web server performance, and varying business choices, such as the breadth of your company's advertising campaign, can determine the number of user requests your Web sites will be required to handle. You should have a good idea of what those loads will be and simulate them on your servers before you put them on line. These are a few of the reasons why there is no silver bullet recommendation on how to tune your Web server.

Performance tuning a Web server should be viewed as an art as much as a science: trial and error can be an important technique in determining what settings and hardware work best for your Web site's requirements. While it's crucial that you understand the technical settings discussed in this document, it is equally important that you understand the profile of your applications or Web sites and how they behave under different conditions. Like a painter who sketches in charcoal to develop a sense of how he wants to create a painting, you need to have a plan for evaluating your Web server performance. The first step is to set up a controlled environment in which to test your Web site, conduct performance analysis of predicted loads, and then measure performance in that environment before you expose your Web server to the Internet. Since the performance of the server can vary greatly with the amount of browser traffic hitting your site during different periods, be sure to monitor your test site under a number of different loads to capture a true picture of activity on the server. During this period, you can develop backup plans to help prevent your site from going down due to any problems during or after deployment.

To improve server performance, examine every part of the system for potential bottlenecks. Bottlenecks can be caused by inadequate or improperly configured hardware or by software settings in either IIS or Windows 2000. A good monitoring plan checks performance in all areas.

Once you know how your server is performing, you can begin to make changes aimed at improving performance. Changes should be made one at a time, with a tested rollback plan, or it becomes difficult to assess the impact of individual changes.

After each change is made, continue to monitor to see if the change has had the desired effect. If unwanted side effects are observed, the rollback procedure can be performed, returning the server to its previous state. Because changes to one resource can cause bottlenecks to appear in other areas, it is important to check on the performance of all resources after you make a change. Once you have assessed the impact of a change, you can determine whether further changes are necessary.