I can recreate the failure condition by manually running the same things on a bare-trusty worker from the command like, so this should be fairly easily reproducible. Also doesn't seem to be an infrastructure-specific issue, so switching projects accordingly.
I can recreate the failure condition by manually running the same things on a bare-trusty worker from the command like, so this should be fairly easily reproducible. Also doesn't seem to be an infrastructure- specific issue, so switching projects accordingly.