Mismatch on http_req_failed... count on prometheus

Grafana k6 k6-browser
1

Hi Team,

I’m using k6 with Prometheus. And recently i noticed Grafana stars to show multiple records (20 records) for a single failure.

When i try to troubleshoot, looks like the single failures insert multiple records on the measurement table k6_http_req_failed_rate with same time interval (15s scrape interval).

image
image1410×899 118 KB

And any idea how to fix this.

3

Hi @gerardlj

Just to double-check, do you use the Prometheus remote write | Grafana k6 documentation output?

Does this duplication presented each time when failure occurs?

4

Yes, I’m using the prometheus remote write. Also the duplication observed as of now on http_req_failed metrics only, other metrics i didn’t investigate.

Observed when a valid failure happens it show multiple records on the Grafana.

5

Apologies for the delayed response, but it sounds like a bug. Could you please open an issue GitHub · Where software is built and provide as many details as possible.

Thanks!