Using Performance Monitor

As distributed applications have matured, performance analysis has become increasingly important. Maximizing hardware and software performance is now a priority. The goal of performance analysis is to determine which of the three major parts of the server — memory, disk I/O subsystem, or CPU — is the bottleneck.

Performance analysis is about collecting empirical data as your application runs in its intended network infrastructure. Such data monitoring can show how your application uses the existing hardware resources. You can then identify the potential resource constraints that might reduce the performance of your application.

If you're new to performance monitoring, it is important to realize that detecting performance bottlenecks is part art, part science, and generally gets better with experience. For example, you may think that memory is the constraint and go out and buy more memory for your server — only to find out that your application still isn't performing because the CPU is the real problem.

One of the best tools for doing performance analysis is the Windows NT Performance Monitor (also called PerfMon). With Performance Monitor, you can easily analyze the workload use of the following hardware and operating system resources.

• Memory
• Disk I/O
• CPU
• Objects
• Threads
• Processes

With Performance Monitor, you can identify hardware resource constraints, component deployment problems, queue congestion, and countless other performance-related views of your application.

For More Information   For more information on using Performance Monitor, see Performance Monitor Help. You can run Performance Monitor from the Administrative Tools menu of Windows NT.