Stale NFS file handle causes Error / Abort

[Tobias Baldauf]

Hi again!

Déjà Dup has started dying on me once more.

Here's my setup:
- Déjà Dup 8.1
- Duplicity 0.5.15-0intrepid1
- Ubuntu 8.10 up-to-date

I attached the relevant part of Déjà Dup's logfile.

The problem arises with a stale NFS file handle on one of my files. I understand that a stale NFS file handle is a problem outside Déjà Dup's responsibility, but I feel that a great backup-utility like Déjà Dup should not abort the whole backup-process because of it.

It would be great if Déjà Dup would continue working and display a logfile of some sort to the user after finishing the backup, notifying him/her that certain files could not be backupped because of e.g. a stale NFS file handle or whatever the reason.

Right now, Déjà Dup simply exists with an "unknown error" and breaks off the backup process. But leaving a great many files unsecured and exiting without a helpful message like "Hey, there a stale NFS file handle on file /foo/bar and that's why I had to quit!" is not good enough for a great app like Déjà Dup which I really came to enjoy!

Please let me know if I can assist any further on squishing this bug!

[Michael Terry (mterry)]

Nowadays, deja-dup will show the exception, so that may give a hint. But I agree it should behave better -- ideally continuing. This is similar to the follow-on duplicity work from bug 364187 (to skip files that can't be read/backed-up, but still notify user about it).

[Michael Terry (mterry)]

Marking invalid on the DD side, since it can't do anything more about this than report the error. Proper fix will come in duplicity.

[Launchpad Janitor (janitor)]

[Expired for Duplicity because there has been no activity for 60 days.]