If ccnet has been configured to not desync when a repository is unavailable (allow_invalid_worktree), and the repository is unavailable (not mounted, for example), the seafile.log file is filled with a warning, repeating once a second:
[11/02/14 16:05:34] repo-mgr.c(121): Failed to access worktree /mnt/photos_usb/Testing for repo 'Testing'(cd903994)
[11/02/14 16:05:35] repo-mgr.c(121): Failed to access worktree /mnt/photos_usb/Testing for repo 'Testing'(cd903994)
[11/02/14 16:05:36] repo-mgr.c(121): Failed to access worktree /mnt/photos_usb/Testing for repo 'Testing'(cd903994)
[11/02/14 16:05:37] repo-mgr.c(121): Failed to access worktree /mnt/photos_usb/Testing for repo 'Testing'(cd903994)
If a repository is stored on removable media and often unavailable, this can cause a large log file to be created. It might be better if this warning was given once (when the worktree first became unavailable), instead of once per call to seaf_repo_check_worktree().
If ccnet has been configured to not desync when a repository is unavailable (
allow_invalid_worktree
), and the repository is unavailable (not mounted, for example), the seafile.log file is filled with a warning, repeating once a second:If a repository is stored on removable media and often unavailable, this can cause a large log file to be created. It might be better if this warning was given once (when the worktree first became unavailable), instead of once per call to
seaf_repo_check_worktree()
.