fix: k6-prometheus-dashboard: Add a Max reducer function when calculating request total

[monteiro-renato]

Hey, I've noticed a bug in the dashboard when displaying the total http requests.

The logs were showing a considerably higher amount of requests than the aggregation displayed in the Grafana dashboard.

     http_reqs......................: 2291   7.593854/s
       { name:create dialog }.......: 2287   7.580595/s
     http_reqs......................: 2346   7.757946/s
       { name:create dialog }.......: 2342   7.744718/s
     http_reqs......................: 2436   8.046441/s
       { name:create dialog }.......: 2432   8.033228/s
     http_reqs......................: 2490   8.248169/s
       { name:create dialog }.......: 2486   8.234919/s
     http_reqs......................: 2373   7.832228/s
       { name:create dialog }.......: 2369   7.819025/s

The aggregation displayed in Grafana should be 2287 + 2342 + 2432 + 2486 + 2369 = 11916

However, we were only seeing a total of 7141 requests.

Basically, if we queried over 13:49:30, the values of the pods no longer pushing metrics were not being taken into consideration.

[codebien]

Hey @monteiro-renato, unfortunately this dashboard isn't designed to support distributed cases. If you need it, then you have to maintain it on your side.

This is expected to be a starting point for major users where they have to extend based on their cases. I close the pull request, feel free to open a new one on k6-core repository if you think this is needed.