Hello, i’m struggling to understand metric called percentile, tried googling it and so on. But it is hard to grasp what are those timings we get with api tests. As I though at first that i.e. 90th percentile metric shows that 10% of API tests, have 1.935 duration, but I am not sure anymore. And if I am right, how does it actually get the percentile ?
It’s 90th percentile. It means that 90% of the requests should be faster than given latency, in your case 1.935 seconds I believe. In other words only 10% of the requests are allowed to be slower than this time.
2 Likes
A small correction to
It means that 90% of the requests should be faster than given latency, in your case 1.935 seconds I believe. In other words only 10% of the requests are allowed to be slower than this time.
In this context(k6 showing some percentiles) it means that 90% of the requests were not should.
If this was about k6 thresholds using percentiles - then yes it will be about what is allowed for the test to pass.
And if I am right, how does it actually get the percentile ?
Basically what it says in Percentile - Wikipedia
1 Like