testing - How to interpret graph result of jmeter load test?


Keywords:testing 


Question: 

I am running a test with jmeter on osx against a server cluster. Now I am wondering how to interpret the test results taken with "graph result".

This is an example output running 50 threads on a macbook pro 2015:

enter image description here

It states 2,090.949/minute throughput. Is this 2M requests per minute? That would equal 33K/s ?! The y-Axis indicates ms, is this latency or time for each request? What do the numbers Average, Median etc. indicate, are those ms per request?

Thank you for any help on this.


2 Answers: 

  • Throughput - number of requests processed by the server per unit time. In your case, the server was able to process 2090.949 requests in a minute. (It is not 2 million)
  • Response Time - y-Axis indicates the response time - the time between once the request is sent from Jmeter till the response received by the Jmeter. It also includes any latency.
  • Average - a simple average of all the response times
  • Median - It is the middle number from the sorted list of numbers.

Lets assume I send 11 requests to access www.google.com. The response times for those 11 requests were 1,1,2,1,3,1,4,5,1,41,28 seconds.

If i find a average 88/11 = 8 seconds. If i find a median = 1,1,1,1,1,2,3,4,5,28,41 = 2 seconds

Average here shows that it takes 8 seconds to get the response from google. But actually it is not true as the average has the disadvantage of being affected by any single value being too high!!

That is where median comes into picture. It shows here that 50% of the requests got their response within 2 seconds.

 

It is not 2M requests per minute. It is 2090 requests per minute.

The median is the central point of a data set.

Average is a synonym for arithmetic mean - which is the value obtained by dividing the sum of a set of quantities by the number of quantities in the set.