Post Run: Cache upload failed because file read failed with EBADF #332
Description
I'm using a self-hosted GitHub Runner, and frequently I receive the following error during the Post Run actions/setup-node@2:
Post job cleanup.
/Users/davidwinter/actions-runner/_work/_tool/node/16.6.2/x64/bin/npm config get cache
/Users/davidwinter/.npm
/usr/bin/tar --posix --use-compress-program zstd -T0 -cf cache.tzst -P -C /Users/davidwinter/actions-runner/_work/nimblenote/nimblenote --files-from manifest.txt
Warning: uploadChunk (start: 33554432, end: 67108863) failed: Cache service responded with 503
/Users/davidwinter/actions-runner/_work/_actions/actions/setup-node/v2/dist/cache-save/index.js:3682
throw new Error(`Cache upload failed because file read failed with ${error.message}`);
^
Error: Cache upload failed because file read failed with EBADF: bad file descriptor, read
at ReadStream.<anonymous> (/Users/davidwinter/actions-runner/_work/_actions/actions/setup-node/v2/dist/cache-save/index.js:3682:31)
at ReadStream.emit (events.js:210:5)
at internal/fs/streams.js:167:12
at FSReqCallback.wrapper [as oncomplete] (fs.js:470:5)
Sometimes, re-running the job will succeed fine and other times it will fail again. It's making my test run success rates completely unreliable and random.
It is occurring on both a macOS and Linux runner.
I've seen some places about changing ulimit -n but would appreciate any insight into that fix if it is known?