I am running my Jmx scripts on stage server. Before I could find out benchmark it breaks the (server goes down) server. I have to restart it and then start again.

  1. Is their a way, I can monitor and stop the scripts manually before it breaks the server?
  2. Also in what terms/keywords/jmeter-jargons I can answer to client that your server can sustain this much load and beyond that it would break?

My Laptop info:

   Operating System: Windows 7 Professional 64-bit (6.1, Build 7601) Service Pack 1 (7601.win7sp1_gdr.150202-1526)
           Language: English (Regional Setting: English)
               BIOS: BIOS Date: 08/18/15 02:15:18 Ver: A07.00 
          Processor: Intel(R) Core(TM) i5-5200U CPU @ 2.20GHz (4 CPUs), ~2.2GHz
             Memory: 8192MB RAM
Available OS Memory: 8096MB RAM
          Page File: 9754MB used, 6436MB available
        Windows Dir: C:\Windows
    DirectX Version: DirectX 11
DX Setup Parameters: Not found
   User DPI Setting: Using System DPI
 System DPI Setting: 96 DPI (100 percent)
    DWM DPI Scaling: Disabled
     DxDiag Version: 6.01.7601.17514 32bit Unicode
  • Are you executing your script on the server hosting the service/website ?
    – Florent B.
    Commented Oct 25, 2016 at 18:31
  • Ok so to clarify, your load test takes down your service so fast that you cant even see the results? Isn't that enough to raise a major concern?
    – ECiurleo
    Commented Nov 1, 2016 at 15:06

1 Answer 1

  1. Don't run application under test and JMeter on the same machine to avoid the mutual interference, JMeter load test is quite "heavy" operation itself and in case of hardware resources lack you won't be able to tell for sure what processes are to blame.
  2. Don't try to overwhelm the server from the very beginning, start with simulating anticipated number of real users. See Use the correct Number of Threads chapter for detailed explanation
  3. Increase your load gradually. If you kick off all the virtual users at the same time you may run into the situation when your application stops responding and you don't know what is the maximum amount of virtual users it may handle while providing reasonable response time and what are the trends. So consider using reasonable ramp-up period. As per documentation:

    Ramp-up needs to be long enough to avoid too large a work-load at the start of a test, and short enough that the last threads start running before the first ones finish (unless one wants that to happen).

    Start with Ramp-up = number of threads and adjust up or down as needed.

  4. Don't forget about simulating "think times". Real users don't hammer the server non-stop, they need some time to "think" between operations. Use JMeter Timers to mimic it.

  5. Always monitor baseline OS metrics i.e. usage of CPU, Ram, paging file, network bandwidth and disk input/output counters. See How to Monitor Your Server Health & Performance During a JMeter Load Test to learn how you can measure the impact of your load test on the server hardware components.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.