Hi,I am encountering a big performance problem with the <ig:gridView> as compared to the <h:dataTable>, or to the <t:dataTable>.We have used JProfiler to measure the response time of a request to a JSP which renders 15 rows of a simnple POJO, including passthrough a service layer (EJB), and a data layer (hibernate). The results were as follows:<t:dataTable> - 900 Milliseconds<ig:gridView> - 2000 MillisecondsHow to reproduce:
1). Create a JSF managed bean with a getter method to a List which holds 15 POJOs. 2). Create a JSP file with a <t:dataTable> which renders the 15 POJOs as rows. 3). Create a JSP file with a <ig:gridView> which renders the 15 POJOs as rows. 4). Compare the 2 pages' response times.
Hello Danny,
I hope the behavior which you are getting due to the additional features provided by Infragistics. Its gives you a better look and feel with inbuild AJAX support which may causing some extra response time. Could you provide me some steps to use the JProfiler to measure the response time?
Thank you
Hello roshan,
Thank you for your reply.
The problem I am encountering is with the rendering phase of the <ig:gridView>.
I have created 2 jsp pages which both contain AJAX support. One with the <t:dataTable> tag and the <a4j> AJAX javascript library. And the other with the <ig:gridView> and the Infragistics AJAX javascript libraries.
The difference, in response time, between the two pages, was really big in rendering the response.
Steps to use JProfiler
1). Install JProfiler, with a valid license.
2). Start JProfiler, and configure it to work with your application server's execution script (JBoss, WebLogic, etc...)
3). Open a browser and point it to the first JSP file with the <t:dataTable> example table.
4). Go to JProfiler and start to record CPU data.
5). refresh the JSP page
6). Go bakc to JProfiler and stop the recording of CPU data.
7). save the results to a file.
8). Repeat steps 3-7 with the second JSP page (ig:gridView).
9). compare the results of the two JProfiler CPU snapshots