Infinite receiving

[Milo123459]

https://github.com/railwayapp/nixpacks/actions/runs/3073248385/jobs/4965261535#step:5:15564

Please let me know if you need more context.

[Swatinem]

Is this reproducible? I think this is rather an upstream problem. The whole cache code itself is just using upstream.

[Milo123459]

Not sure, it just happened really.

[LesnyRumcajs]

Hey @Swatinem, seems we encountered a similar issue https://github.com/ChainSafe/forest/actions/runs/3246430665/jobs/5325159334

Do you have any pointers except for timing out early?

[ivankovnatsky]

We experiencing this quite often, I would say roughly 10% of all builds.

[Swatinem]

I’m pretty sure that is an upstream issue. This action delegates the whole caching step itself to the upstream @actions/cache package. Issues go here: https://github.com/actions/toolkit/issues

[ivankovnatsky]

Let's close this one then?

[abonander]

I don't see a relevant issue on actions/toolkit but this definitely still happens to us periodically, it's like the download bandwidth just slows to zero and stops, but it doesn't detect this and time out or try to recover: https://github.com/launchbadge/sqlx/actions/runs/6654941857/job/18355715032?pr=2545