hundreds-father-404
04/07/2021, 8:44 PMException: String("Cannot strip prefix __output from root directory (Digest with hash Fingerprint<bb11eda082823e90c95f959f8b89c6886423521773bb24bb31e8c25454d0740b>) - root directory didn\'t contain a directory named __output but did contain directory named: shellcheck-v0.7.1")
https://github.com/pantsbuild/pants/runs/2291026549#step:9:234
The confusing thing to me is why it's flaky22:01:06.28 [INFO] input snapshot files: ('shellcheck-v0.7.1.linux.x86_64.tar.xz',)
185
22:01:06.28 [INFO] input snapshot dirs: ('__output',)
186
22:01:06.28 [INFO] argv: ('/usr/bin/tar', 'xf', 'shellcheck-v0.7.1.linux.x86_64.tar.xz', '-C', '__output')
22:01:06.51 [INFO] output files: ('shellcheck-v0.7.1/LICENSE.txt', 'shellcheck-v0.7.1/README.txt', 'shellcheck-v0.7.1/shellcheck')
197
22:01:06.51 [INFO] output dirs: ('shellcheck-v0.7.1',)
The output files should have __output
in the prefix, and we tell the Process
that output_dirs=("__output",)
, so presumably Tar is behaving correctly and writing there
Using GNU Tar 1.30, which is very modern and supports the -C
flagenough-analyst-54434
04/09/2021, 12:43 AMhundreds-father-404
04/09/2021, 12:54 AMlint ::
30 times with no local cache, I can't get it to fail once. I still can't get it to fail on macOS and Linux https://github.com/pantsbuild/pants/pull/11856
I'm running out of ideas (and am not going to think about this until Monday). But any thoughts on figuring out why this would stop flaking in that PR? My idea is to revert some of the changes like --no-pantsd
and see if one of them is causing it