Win a copy of Design for the Mind this week in the Design forum!
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic

Stress Testing multiple urls on JMeter

 
Nelo Angelo
Ranch Hand
Posts: 44
Chrome Eclipse IDE Java
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Hello everyone,

I am trying to stress test multiple URLs using JMeter. The URLs are contained in a CSV file. I am not able to use the CSV Data Config set properly. Any suggestions on how to do the stress test?
 
sugan san
Ranch Hand
Posts: 47
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
You should use CSVdataconfig element, Please see the below Url
http://www.technix.in/jmeter-problems-and-solutions/
 
Andrew Monkhouse
author and jackaroo
Marshal Commander
Pie
Posts: 11865
194
C++ Firefox Browser IntelliJ IDE Java Mac Oracle
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
I think that this is a classic example of The XY problem. A year ago the original poster asked about stress testing using a CSV file. This is unlikely to produce a real stress test, although it may provide a Smoke Test.

Consider a couple of cases:

  • The user creates a JMeter script that goes to every URL in the CSV file. What have they stressed?

    If the application being accessed has only just started up, then there may be the illusion of stress, as various classes are loaded into memory, initializations are performed, connections are opened to other services (or databases), ... But none of these are indicative of real-life performance.

    Even if the application has been running for some time, running through a list of URLs are unlikely to provide a real-life scenario. Consider testing forum software - if you are lucky, you have 95% anonymous viewers, 4% logged in users, and 1% actively posting. (A more complex scenario would be easy to devise as well). How are you going to simulate that using a set of URLs in a CSV file?

  • The user creates a JMeter script that goes to every URL in the CSV file multiple times (presumably set by a value in the CSV file).

    We are still not stressing anything in a real-life scenario. Let's imagine that I request this topic 10 times. The first time through the record(s) get pulled from the database, and the topic is produced. Every time after that, you will start hitting some form of caching - whether it is at the web server, the application server, the code, or the database. It is not inconceivable that I get slightly better results for each of those 10 requests - and I will have learned nothing.



So the first question for Nelo is: what are you hoping to achieve? If it is a true stress test, then you will probably find it easier to build a test script from scratch than to try and build one around a CSV file. If all you want is a smoke test, then either of the above approaches will work, although be prepared that the site Sugan San points you to seems to be incomplete (at best) and straight out wrong at times.
 
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic