Topic: More workstation backup issues

[DanB35]

After discovering that the workstation backup did not, in fact, fail silently as I'd thought it did, I'm still having some trouble getting it to work regularly.  I'm running the backup (or trying to) from my SME 8.1 server to a CIFS share on a FreeNAS 9.2.1.6-BETA server which (share) is only used for this backup.  For a while, the backup script wasn't able to mount the share at all unless I'd recently restarted the CIFS service on the FreeNAS box.  Now, however, it won't run at all.  When I try to run the backup manually, I get this:

Code: [Select]

[root@e-smith dan]# /sbin/e-smith/do_backupwk
Aborting program. User refused to continue while asking: The last file of the set is not present in /mnt/smb/e-smith.familybrown.org/set2 , please provide it.
Error while running dar: 4 at /etc/e-smith/events/actions/workstation-backup-dar line 505.
Backup terminated: backup failed - status: 7424

It looks to me like the catalog and the actual backup file set don't match.  I don't have a problem with deleting the backup file set if necessary, but I don't want to overly confuse the backup system, which appears to be a bit confused already.

[Stefano]

I can't help you with your issue but, since you are using Freenas, I strongly suggest you to use NFS share as it is at least 3 times faster than cifs/smb one

HTH

[DanB35]

I'd tried to set up the backup via NFS previously, and it hadn't worked, but I didn't recall the exact errors.  So I figured I'd set it up again to see what the errors were, and post them here.  Set up a new dataset on the FreeNAS server, a new NFS export, and gave it a try.  Of course, the Technician Proximity Effect is in full force--it worked fine, with no errors.

I'll let this run for a few days--if it runs reliably, that'll be good enough for me.  Still not sure what's going on with the error I posted about, though.

[TerryF]

It looks to me like the catalog and the actual backup file set don't match.  I don't have a problem with deleting the backup file set if necessary, but I don't want to overly confuse the backup system, which appears to be a bit confused already.

I have found that was only way to get system back on track, or simply rename the directory on the backup device so SME creates a completely new set etc using the old settings and see what occurs, that way you still have all the original files etc.or create a completely new backup job.

Fingers crossed  for good luck

[DanB35]

Going into a new directory, it worked fine for two manually-initiated backups, one full and one incremental.  I'll have to see if it runs reliably on schedule, though I'd expect it would.

[Drifting]

I have exactly the same problem. Also using Freenas (Which I use all over for a backup destination)

Can't remove directory /mnt/smb/dc01.dc.local/set2 (Directory not empty) at /etc/e-smith/events/actions/workstation-backup-dar line 575
Aborting program. User refused to continue while asking: The last file of the set is not present in /mnt/smb/dc01.dc.local/set1 , please provide it.
Failed to add set /mnt/smb/dc01.dc.local/set1/full-201408020001 to catalog.
Backup terminated: backup failed - status: 7424

And

==================================
DAILY BACKUP TO WORKSTATION REPORT
==================================
Backup of dc01.dc.local started at Fri Aug  8 00:01:04 2014
Backup of mysql databases has been done
Mounting backup shared directory <10.0.0.5:backup>
Using set number 2 of 4
Attempt full backup
Backup temp directory /mnt/smb/tmp_dir/dc01.dc.local is mounted and is writable
Backup base file name is full-201408080001
Starting the backup with a timeout of : 28770 seconds


--------------------------------------------
168708 inode(s) saved
  including 26 hard link(s) treated
0 inode(s) changed at the moment of the backup and could not be saved properly
0 byte(s) have been wasted in the archive to resave changing files
0 inode(s) not saved (no inode/file change)
0 inode(s) failed to be saved (filesystem error)
256 inode(s) ignored (excluded by filters)
0 inode(s) recorded as deleted from reference backup
--------------------------------------------
Total number of inode(s) considered: 168964
--------------------------------------------
EA saved for 170 inode(s)
--------------------------------------------
Moving backup files to target directory /mnt/smb/dc01.dc.local/set2
Updating catalog
*** No backup allowed or error during backup ***
Failed to add set /mnt/smb/dc01.dc.local/set1/full-201408020001 to catalog.

I have tried numerous times to recreate the job, deleting all the previous backups (Well I actually renamed them)  The back then runs for a few days and does it again! At a total loss with this. Odd part is I have a number of others servers working fine with the same setup.

Paul

[janet]

DanB35 & Drifting

Use process of elimination.
Run the backup to a locally connected USB drive (ie physically attached to sme server).
Let it run for a few weeks to prove stability etc.

Possibly your issues are network related, it could even be erratic or faulty NICs or a hub/switch going to sleep or faulty/poor cabling & connections, or the NAS going to sleep (or similar low power shutdown mode etc), something is disrupting the backup process & it's not finishing correctly, so you see the errors that get created next time you try to do a backup. Possibly the mount is not being unmounted properly, so check this manually between backup sessions.

Google everything ie

Backup terminated: backup failed - status: 7424

There are plenty of bugs & posts with similar errors.
eg
http://bugs.contribs.org/show_bug.cgi?id=6933#c1

[DanB35]

I'm sorry, I should have reported back here much sooner.  Now that I have it set up to back up to the NFS share, it's been running smoothly for about the last 6 weeks.  I have no idea if that's due to NFS vs. CIFS or something else.

[janet]

DanB35

Thanks Dan, your result is suggestive to avoid cifs where possible.

[Drifting]

Mmm.

It is not possible for me to try USB, these servers I support free of charge as they are for a charity, and are up the other end of the country. Looks like I shall have to give NFS a try. Odd when I have a number of other identical setups all working fine, yet another of life little mysteries.

Paul

[janet]

Drifting

....are up the other end of the country.

It can all be done by ssh remote control.
Just get a a local user to purchase a USB drive (relatively cheap these days) & connect it physically.
You then connect remotely via ssh, test the drive location (fdisk -l), format it ext3/4 & partition it & so on (follow the available Howtos), mount & test manually & unmount, then you are ready to configure & test automated backups.

[Drifting]

Thanks Janet

That information was really helpful, never occurred to me. (Slaps head)
I think the part about this that bothered me, is that I have 4 other servers (same charity and setup) that are working fine? Hate problems like this, never make any sense.

Paul

[TerryF]

Thanks Janet
That information was really helpful, never occurred to me. (Slaps head)

Just goes to show you can always learn something new..