Received data differs from each other

Grafana k6 OSS Support
1

Hello, I’m sending the same request with K6 and receive data was really differs from each other.
The issue is connected with “Iterations count”
For the first request i get the iterations count as 3500 and after the second request (1sec later) I see the iterations count 8900. The difference was really big, so I want to understand what can matter to following issue.

The response bodes see above

  1. ✓ status is OK

    checks…: 100.00% ✓ 3551 ✗ 0
    data_received…: 6.4 MB 128 kB/s
    data_sent…: 2.3 MB 47 kB/s
    grpc_req_duration…: avg=165.86ms min=14.37ms med=130.42ms max=1.7s p(90)=236.11ms p(95)=305.44ms
    iteration_duration…: avg=284.86ms min=89.31ms med=243.47ms max=1.86s p(90)=376.89ms p(95)=502.23ms
    iterations…: 3551 71.005159/s
    vus…: 100 min=100 max=100
    vus_max…: 100 min=100 max=100

  2. ✓ status is OK

    checks…: 100.00% ✓ 8966 ✗ 0
    data_received…: 16 MB 787 kB/s
    data_sent…: 5.8 MB 286 kB/s
    grpc_req_duration…: avg=112.64ms min=15.39ms med=110.23ms max=283.36ms p(90)=159.93ms p(95)=179.07ms
    iteration_duration…: avg=224.34ms min=85.26ms med=219.9ms max=421.3ms p(90)=276.6ms p(95)=300.01ms
    iterations…: 8966 443.620645/s
    vus…: 100 min=100 max=100
    vus_max…: 100 min

5

Hi @Gohar,
welcome to the community forum :wave:

The difference was really big, so I want to understand what can matter to following issue

Your service under test on the 2nd run is faster, as you can see from the reported p95 metric. Does the service implement some caching?

Do you see the same behaviour if you make a few requests? e.g. only 1-10 requests
Is the waiting time between the runs relevant? What does it happen if you wait 1 minute instead of 1 second?

It isn’t easy to investigate what is happening without checking your service’s metrics. Did you check them?

1 Like