We are currently seeing outage of the caching on the side of Github (see issue: https://github.com/actions/cache/issues/820). While this is something they have to fix, I think this action should be fault-tolerant when cache is unavailable, which means refetching and reinstalling all dependencies. Right now the action is failing with: getCacheEntry failed: Cache service responded with 503. This has been the case before (related: https://github.com/actions/cache/issues/698).
One of the proposed solutions to make sure cache never gets in the way of building and deploying again, was making all setup-{} actions on the side of Github more fault tolerant (https://github.com/actions/cache/issues/698#issuecomment-992565426). I think this should be implemented in this action as well.
We are currently seeing outage of the caching on the side of Github (see issue: https://github.com/actions/cache/issues/820). While this is something they have to fix, I think this action should be fault-tolerant when cache is unavailable, which means refetching and reinstalling all dependencies. Right now the action is failing with:
getCacheEntry failed: Cache service responded with 503
. This has been the case before (related: https://github.com/actions/cache/issues/698).One of the proposed solutions to make sure cache never gets in the way of building and deploying again, was making all
setup-{}
actions on the side of Github more fault tolerant (https://github.com/actions/cache/issues/698#issuecomment-992565426). I think this should be implemented in this action as well.