Time Machine no longer works after 8.3.0 upgrade.

[witchbutter]

Upgrading from 8.3.0 RC1 to RELEASE all of my time machine shares no longer work. I can browse to them in finder, I can see my sparsebundles and view their contents, I can go into time machine if I choose "browse other time machine disks" however all 3 macs I have no longer can back up. Time machine says the "The backup disk is not available"

 

[bollar]

I assume you've done the obvious things -- verified that Time Machine share is setup correctly in AFP; turned the AFP service off and on; maybe even rebooted the FreeNAS?

You might also want to look at the Mac Console to see if the error on that side gives you more info. Time Machine Buddy can make this easier and also check for faulty permissions and some other problems that might be the cause.

 

[witchbutter]

I assume you've done the obvious things -- verified that Time Machine share is setup correctly in AFP; turned the AFP service off and on; maybe even rebooted the FreeNAS?

Yes I did.

The console output is cryptic to me in this case:

Code:

11/17/12 11:20:50.846 AM com.apple.backupd[613]: Starting automatic backup
11/17/12 11:20:50.902 AM com.apple.backupd[613]: Backup failed with error: 19
11/17/12 12:21:21.083 PM com.apple.backupd[697]: Starting automatic backup
11/17/12 12:21:21.102 PM com.apple.backupd[697]: Backup failed with error: 19
11/17/12 1:03:59.462 PM com.apple.backupd[310]: Attempting to mount network destination URL: afp://witchbutter@hlin._afpovertcp._tcp.local/wintermoon
11/17/12 1:04:29.482 PM com.apple.backupd[310]: NAConnectToServerSync failed with error: 64 for url: afp://witchbutter@hlin._afpovertcp._tcp.local/wintermoon
11/17/12 1:05:29.505 PM com.apple.backupd[310]: Failed to resolve network service using name = hlin type = _afpovertcp._tcp domain = local.
11/17/12 2:01:26.093 PM com.apple.backupd[478]: Starting automatic backup
11/17/12 2:01:26.159 PM com.apple.backupd[478]: Backup failed with error: 19
11/17/12 3:01:56.403 PM com.apple.backupd[576]: Starting automatic backup
11/17/12 3:01:56.873 PM com.apple.backupd[576]: Backup failed with error: 19
11/17/12 3:24:54.599 PM com.apple.backupd[725]: Starting manual backup
11/17/12 3:24:54.628 PM com.apple.backupd[725]: Backup failed with error: 19
11/17/12 3:29:08.784 PM com.apple.backupd[788]: Starting manual backup
11/17/12 3:29:08.786 PM com.apple.backupd[788]: Attempting to mount Time Machine network destination configured prior to 10.5.5 - alias does not contain an embedded URL
11/17/12 3:29:08.797 PM com.apple.backupd[788]: Network destination already mounted at /Volumes/wintermoon-1 but mount point not owned by root - attempting to remount as root
11/17/12 3:29:08.810 PM com.apple.backupd[788]: Attempting to mount network destination URL: afp://witchbutter@hlin/wintermoon
11/17/12 3:29:08.997 PM com.apple.backupd[788]: NAConnectToServerSync failed with error: 80 for url: afp://witchbutter@hlin/wintermoon
11/17/12 3:29:09.002 PM com.apple.backupd[788]: Backup failed with error: 19
11/17/12 3:29:37.934 PM com.apple.backupd[788]: Starting manual backup

For some reason when the freenas server appears in Finder, it always shows me the SMB shares first. In order to get to the afp connections I have to use Finder > Go > Connect to Server > afp://hlin .
(Hlin is the name of my freenas)
Also I now notice that the destination afp volume has been renamed in /Volumes/wintermoon-1 (this is the old name with -1 appended). Is it storing a history of AFP shares that I need to clear out?

 

[JaimieV]

No issues with TM here; that's in what was 8.3.0rc1 then updated to 8.3.0-RELEASE-p4.

 

[fracai]

The "-1" suffix means there's an existing "wintermoon" folder in the /Volumes directory where the Mac mounts the share. This can happen if the share wasn't properly ejected the last time you mounted it, but OS X also does this whenever it backs up with TimeMachine. I think the OS likes to "own" the share it's using for the backup, so it mounts the share a second time and hides this mount from the UI. That's the meaning of: "Network destination already mounted at /Volumes/wintermoon-1 but mount point not owned by root - attempting to remount as root".

Basically, you should be able to ignore that part, but to rule it out, you could manually unmount all your NAS shares, reboot, and then look at /Volumes from the command line with "ls -la /Volumes". Then try the backup again.

 

[witchbutter]

It appears either my USB stick or my install became corrupted. 2 days ago I installed a new USB stick with 8.3.0 Release, auto-imported my volumes and manually set all my settings back. When I reset all the AFP shares they immediately began backing up without having to create new sparsebundles.