Reliability Improvements

Applies To: Windows Server 2003, Windows Server 2003 with SP1

IIS 6.0 promotes greater reliability and reduces the chance of defective applications compromising overall server performance and application reliability. Applications that do not function properly can adversely affect any Web server, regardless of the reason for the applications faulty functionality. For example, one application can consume CPU resources, causing other applications to perform slowly. In addition, a Web application that fails can potentially bring down a Web server, including all its sites and applications. Web server administrators must sometimes restart the entire Web server, bringing down all its applications and sites, in an attempt to update or fix one faulty application. IIS 6.0 offers solutions for all these scenarios.

The IIS 6.0 redesigned architecture, which includes two new request processing models, worker process isolation mode and IIS 5.0 isolation mode, helps you avoid the problems caused by faulty applications. Worker process isolation mode, which is the default application isolation mode for IIS 6.0, separates key components of the World Wide Web Publishing Service (WWW service) from the effects of malfunctioning applications by using isolated worker processes. In IIS 5.0 isolation mode, requests are processed similarly to how they were processed in IIS 5.0. This process model is intended for existing applications that do not initially work when you test them during deployment testing in worker process isolation mode. For more information about IIS architecture changes, see IIS 6.0 Architecture.

Worker process isolation mode is the default process model in IIS 6.0 because it enables the greatest number of benefits, many of which are achieved by using application pools. An application pool is a grouping of one or more URLs that are served by a worker process or set of worker processes. You can configure specific settings for each application pool and the worker processes that serve it.

If your server is running applications that contain imperfect code, using worker process isolation mode allows you to isolate a defective application in an application pool. You can then use health monitoring (a feature that makes an application pool self-healing), worker process recycling (periodic restarts of unstable applications or ones that leak memory), and rapid-fail protection (auto-restart of failed applications) to specifically address the applications faulty functionality, which results in fewer or no system restarts. For more information about how IIS improves reliability of applications, see Running IIS 6.0 as an Application Server.