engleski

EMPIRICAL COMPARISON OF DIFFERENT VIEW STATE APPROACHES ON PERFORMANCE OF ASP.NET WEB APPLICATIONS

Web forms are still common choice for development of ASP.NET applications, despite the fact that ASP.NET MVC is becoming increasingly popular. One of the main paradigms of Web forms are postback and ViewState. A situation when Web form issues POST request back to itself is called a postback. ViewStatemechanism was introduced to bridge the gap between postbacks. It provides an illusion of statefulness between two requests on the same Web form, as if the state of Web form hasn’t changed since the last request was finished processing. In order to utilize ViewState no configuration is needed and it is common practice to leave default ViewState settings. This paper shows how ViewState size affects performance of Web form. It shows that ViewState size can easily grow large and that default settings can affect performance by enlarging both responses and requests. Two alternatives are analyzed: sending compressed ViewState to client and leaving ViewState on server.We show that best choice is to leave ViewState on server. Since that method is not always applicable, we show that the second best method is compression. It is applicable in almost all situations and will have good impact on performance.

ASP.NET ; Web forms ; performance ; ViewState ; compression ; Session

nije evidentirano