Whats the difference between 90 percentile and Avg 90 % ?How are they calculated? which value (more appropriate) should we consider while publishing reports?
As there is around 3-4 sec difference between two.
Basically, when removing 10% of worst values:
- 90th percentile, means than 90% of measurements are under the percentile (ex: 90% of users are under 3s)
- Average 90% means the average value of these 90% best values. (ex: Avg 90% = 1.2s when Percentile is 3s)
Percentile is often preferred, as it is performance goal oriented. If your SLA is that the response time should be under 3s, then if the percentile 95 (for example) is under 3s, it shows that 95% of the user have an experience that matches your performance goals.
Note that you should use NeoLoad 5.1 to get improvements on the percentiles.
Yaa , I raised this question after using Neoload 5.1, its great, more informative,much more deeper in terms of response time so had a doubt, as per my SLA the response time of all pages for 90% users should be 5 sec or below it, so which column to refer 90 percentile or Avg 90 %?