[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [rdiff-backup-users] Error 1? Final moment permissions problem on ta
Re: [rdiff-backup-users] Error 1? Final moment permissions problem on target?
Tue, 15 Apr 2014 12:27:18 +0100
Mozilla/5.0 (Windows; U; Windows NT 6.0; en-GB; rv:22.214.171.124) Gecko/20110414 Lightning/1.0b2 Thunderbird/3.1.10
On 15/04/2014 10:38, Claus-Justus Heine wrote:
Could it be that perhaps the UID mapping is not set up correctly? Being
able to do r/w but not being able to preserve times and set permission
could indicate that somehow the NFS-Server "thinks" that the files are
not owned by the user issuing the command. Cam you do a "chmod"
manually? Or can it perhaps be that only the mount-directory
/mnt/new-dest is not owned by the backup user (the chmod error message
of rdiff-backup only mentions the mount-point)?
Useful pointers, thank you. What I find is that all the *directories*
under /mnt/exist-dest and /mnt/new-dest have the same permissions
(exactly as I had intended by using the -a parameter on the cp command
to copy the existing backups to the new machine), but the mount
points, indeed, do differ in owner and group. Both mount points have
permissions 40777 (shown by mc), so that any user can read and write.
In detail, the mount point of the existing backup /mnt/exist-dest has
owner ron, group users, permissions 40777.
The mount point of the new backup /mnt/new-dest has owner root, group
root, permissions 40777.
An obvious next step is to change the ownership of /mnt/new-dest to be
ron, users, with permissions 40777 (and so match the exist-dest).
Apologies for appearing over-cautious but is there any reason why I
should not just go ahead and do that?
If so, I'll then ask rdiff-backup to check the destination which,
hopefully, will revert this backup attempt, and try again with the
backup. Actually, I'll first redo the cp command to try and have it
finish without its error, as well, we need to be able to rely on the
integrity of the backups we already have and if there is something not
set correctly, restores later might go awry.
Claus, thank you for the suggestion about the mount point. (I was
aware that NFS leads to sub-optimal performance but it was simpler to
set up, as you mention.)