How can we know that the K6 report is reliable

When I run a load test by JMeter and K6 on the same system and test scenario:

  • VUs = 5
  • Ramp up = 0s
  • Iterations = 5
  • The results are bellow:
  • K6 shows http_req_duration - p(95)=2.19s
  • JMeter shows 95% Line = 2.394s
    So which result is correct 2.19s (K6) or 2.394s (JMeter)?

Hi @nxthich

Welcome to the community forum :wave:

Do you consistently get lower numbers with JMeter? Do you get the same results increasing the load?

Based on the “low” iterations and load (5), and the 95% percentile, I would expect some statistical deviation that could be 200ms between test executions. The latency in the endpoint can vary and both results are correct.

I’m also unsure what metric you compare with JMeter, as you did not specify. For k6 http_req_duration is the total time for the request. It’s equal to http_req_sending + http_req_waiting + http_req_receiving (i.e. how long did the remote server take to process the request and respond without the initial DNS lookup/connection times).