Accessibility

Table of Contents

  1. Introduction
  2. RIA performance vs. traditional web app performance
  3. Capturing client-server performance
  4. Large request sizes
  5. Large response sizes
  6. Frequent network calls
  7. Badly timed requests
  8. Conclusion

Improving RPC performance in Flex RIAs

Capturing client-server performance

Given these differences, it's important that RIA performance monitoring happens throughout the development process, and is not necessarily isolated to a dedicated testing phase. You can use tools such as ServiceCapture and Charles to quickly spot and diagnose common performance problems. If you open Service Capture and run the code introduced in the previous section, which uses RemoteObject RPC calls and the Data Transfer Object (DTO) pattern, the application produces the following results (see Table 2):

Initial times for sample application

Table 2. Initial times for sample application

Note: ServiceCapture reports the total HTTP request and response size. As multiple AMF requests are automatically grouped into a single HTTP request, the response size shown for getContacts is actually the response size for bothgetContactsand getOrderHistory combined. ServiceCapture indicates this grouping in the "performance" tab when a particular request is selected (see Figure 2).

AMF grouping

Figure 2. AMF grouping

From the data in Table 2, it is possible to observe how many requests have been sent, the size of each request, and the size of each response. This information can be analyzed and potential performance problems can be quickly spotted. The following are common symptoms of a poorly performing RIA along with suggestions on how to solve these problems.

Page 3 of 8
Previous Page
Next Page