In our test we are validating amount of files processed, after processing completed for each file we increase the counter by 1. But in k6 threshold the count shown is incorrect, we could see in trend shown in graph is correct. Attached the screen shot for checks.
This seems like a local test that streams its results to the cloud, i.e. k6 run --out cloud, right? If so, can you tell me which k6 version are you using? And if you aren’t using the latest one (v0.35.0 at the moment), can you try to run the test with it?
Since, in that test type, the thresholds are calculated by k6 itself, not the cloud, and there used to be an old bug that caused this, I want to rule it out. And if the problem persists, there probably is a bug in k6 we’d like to fix with a high priority…