Thread: What are different Performance counters?

The first and foremost question to address when measuring performance of a web application is "What to measure?" This can be answered by following:

• Identifying and recognizing events that already occur naturally within the application. This method places the burden on those setting up measurements to first identify application-specific milestones.
• Adding instrumentation, by inserting simple monitoring calls like those used by the ARM(Application Response Measurement) standard [ARM1], an existing method for measuring distributed applications. In this approach, developers imbed simple instrumentation calls at logical boundaries in the application’s flow, to enable its subsequent measurement [ARM2]. This frees the measurement tool from the need to know anything about the application, but places the burden on application developers to instrument their code by inserting markers at key points

The following 10 points highlight the purpose of measuring performance of any web application -

1. Verifying that an application meets its service level objectives
2. Detecting abnormal application behavior (typically, slow response)
3. Identifying bottlenecks in application components
4. Comparing builds, versions, or releases of an application
5. Comparing two applications that perform comparable functions
6. Verifying that a system will be able to handle the projected traffic
7. Determining how many users a server environment can handle
8. Predicting bottleneck components as workload levels grow
9. Comparing builds, versions, or releases of an application
10. Identifying components that fail after extended use

When it comes to manual testing, we rely on the factors that I have jotted for performance. The counters you are talking about is taken care by Automated Testing using LoadRunner or SilkPerformer, etc. wherein we have to give various conditions to calculate the performace of an application to the last second.