TW
Guess this can be closed?
Closing this, seems to be rather an issue of not having enough RAM or a kernel issue.
The tracebacks seen come from how borg handles opened segment files: it keeps them open for a while in case they are needed again and uses a size-limited LRU cache...
Just wondering why it starts with that: ``` Starting archive consistency check... ``` If not limited by `--archives-only` it should first do a repository check (and a repo index rebuild),...
Hmm, ok. Maybe it is somehow due to logging (the repository check log lines come from the server, the other log lines are produced locally). But usually, they should be...
But assuming it was a full check, it worked with a freshly computed repo index (it scans the on disk files and writes a fresh index with the mapping `id...
``` OSError: raw readinto() returned invalid length 7149846510 (should have been between 0 and 4763418624) ``` I'ld say this is not a borg error, but an error in Python (maybe...
Tagged this as hardware issue. I am not 100% sure though, could be also a weird software bug in OS / FS / something below borg.
Hmm, seems to be the same issue again: https://github.com/borgbackup/borg/issues/6140#issuecomment-1023223029 Maybe do: ``` ls -l repo_dir ``` After fixing any potential hw issue, you could do a `borg check --repair` to...
``` Create a readonly snapshot of '/' in '//borg-backup' ``` Guess that is not from borg, but there seems to be a slash too much?