[Feature]: Controls for reducing trace size

[jpike88]

🚀 Feature Request

We have some long running tests that can't be broken down, which means their generated traces are quite big (in the 100s of MBs). The screenshots are far too frequent to be necessary, we could cut them down to just 30% of what's happening, and have a much better idea. It would be good to fine tune it so only every 3rd screenshot of what's happening is the new frequency insteand, and maybe retain the final 20 or so to ensure the failure is captured accurately.

Example

No response

Motivation

Smaller trace files

[Skn0tt]

Hi! I understand 100s of MBs adds up quickly. Can you elaborate how you're storing them? Do you have lots of consistently failing tests, so the traces are accumulating, or do you also store traces for successful tests so there's lots of wasted space?

[jpike88]

Neither, i deposit traces for failed tests into another GitHub repo so i can host the html report via GitHub pages. Problem is the blobs are much bigger than the GitHub 100mb file limit, but with 50-70% of those screenshots removed it fits under 100mb nicely. I run a script that goes into each blob and removes every nth screenshot from the data folder to achieve this