[Building Sakai] Load test of Sakai

[Berg, A.M.]

Hi all,

I use Jmeter a lot. We have to be careful to define the difference between a thread and a user. For example for our Learning Management System a user clicks a page on average every 20 seconds. For each http request in Jmeter we need to add an extra delay (normally guassian) of the same value. Sometimes we cheat and have less delay and less threads.

Alan

Alan Berg

Senior Developer / Quality Assurance
Group Education and Research Services
Central Computer Services
University of Amsterdam



-----Original Message-----
From: sakai-dev-bounces at collab.sakaiproject.org on behalf of Pete Peterson
Sent: Thu 8/20/2009 19:02
To: Daniel Lind; sakai-dev at collab.sakaiproject.org
Subject: Re: [Building Sakai] Load test of Sakai
 
Hi Daniel,

FYI: There is confluence site for Sakai Performance work, http://confluence.sakaiproject.org/display/PERF/Home. The content is a little dated, but I think you might find some interesting information there and I also encourage you to share your findings and thoughts with the rest of the community on this site. 

Regards,

Pete

> -----Original Message-----
> From: sakai-dev-bounces at collab.sakaiproject.org [mailto:sakai-dev-
> bounces at collab.sakaiproject.org] On Behalf Of Daniel Lind
> Sent: Thursday, August 20, 2009 12:41 AM
> To: sakai-dev at collab.sakaiproject.org
> Subject: [Building Sakai] Load test of Sakai
> 
> Hi!
> 
> We're investigating the limitations of Sakai when it comes to capacity
> of users. We have created some test plans in JMeter where users are
> taking tests consisting of 25 questions (multiple choice, true/false
> and Short answer) in Test&Quizzes. When we run this test we have come
> to these conclusions:
> 
> One Tomcat: Maximum of 430 simultaneous users from two JMeter
> instances.
> Two Tomcat (on the same server): Maximum of 630 users from two JMeter
> instances.
> 
> 
> When I say max it means the response times are way too long to be
> acceptable (connect >2s and response >10s).
> 
> We are running all the test on only one single server with these specs:
> RedHat 5
> Linux kernel 2.6
> MySQL 5.0
> Sakai 2.5.3
> 10GB RAM
> Intel Xeon 2.5GHz Quad-core
> 
> + tuning in MySQL, Tomcat and Sakai
> 
> Have anyone else done these kinds of tests? What are your experiences?
> 
> 
> 
> 
> Kind regards,
> Daniel Lind
> 
> ------------------------------------------
> Daniel Lind
> UMDAC, Umeå universitet
> 901 87 Umeå Telefon: +46 90 786 53 12
> Daniel.Lind at umdac.umu.se
> ------------------------------------------
> 
> _______________________________________________
> sakai-dev mailing list
> sakai-dev at collab.sakaiproject.org
> http://collab.sakaiproject.org/mailman/listinfo/sakai-dev
> 
> TO UNSUBSCRIBE: send email to sakai-dev-
> unsubscribe at collab.sakaiproject.org with a subject of "unsubscribe"
_______________________________________________
sakai-dev mailing list
sakai-dev at collab.sakaiproject.org
http://collab.sakaiproject.org/mailman/listinfo/sakai-dev

TO UNSUBSCRIBE: send email to sakai-dev-unsubscribe at collab.sakaiproject.org with a subject of "unsubscribe"

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/sakai-dev/attachments/20090820/2e5f7e50/attachment.html