Vuser Test Result

We could have increased the possible max vuser per agent by calibrating JVM options in nGrinder 3.2. The following is the result of our internal test. ‘O’ means that the test with the given vuser count per agent was executed successfully for 15 min. X means that the test was automatically stopped by the controller due to the lack of memory.

Groovy

vuser per agent	1000	2000	3000	4000	5000	6000
1k response	O	O	O	O	O	X
10k response	O	O	O	O	O	X
100k response	O	O	O	O	O	X
process/thread count	10/100	10/200	10/300	10/400	10/500	10/600

Jython

vuser per agent	1000	1500	2000	2500	3000	4000
1k response	O	O	O	O	O	XXX
10k response	O	O	O	XOO	O	XXX
100k response	O	O	O	O	O	OO
process/thread count	10/100	10/150	10/200	10/250	10/300	10/400

In summary, when using Groovy as a scripting language, it supports twice vuser count than using Jython. Even when selecting Jython, it still supports 5 times more vuser count than previous version.

Agent Machine Spec

Agent machine specs we used for this test are

• 2 Core, 4Gram VM (with 3.4Gram free)
• CentOS 5.3 64bit
• JDK 1.6u33 64bit
• nGrinder 3.2 agent

Reponse generating code

The following JSP code is used to simulate the responses. These generate 1k / 10k / 100k responses respectively.

1k response.

<%
Thread.sleep(500);
for (int i = 1; i < 20;i++) {
    out.println("Hello!!! My Name is JunHo Yoon. Welcome! Perf Test!");
}
%>

10k response.

<%
Thread.sleep(500);
for (int i = 1; i < 200;i++) {
    out.println("Hello!!! My Name is JunHo Yoon. Welcome! Perf Test!");
}
%>

100k response.

<%
Thread.sleep(500);
for (int i = 1; i < 2000;i++) {
    out.println("Hello!!! My Name is JunHo Yoon. Welcome! Perf Test!");
}
%>