Deze link vond ik over sparsebundle;
http://discussions.apple.com/thread.jspa?threadID=1237059
ff checken
Re: Sparse bundle - resource temporarily unavailable - how to fix?
Posted: Jun 11, 2008 1:44 PM in response to: Gordon Buchanan
Just went through this myself yesterday. Fortunately, I have a solution of sorts.
THE SHORT VERSION
If your Time Machine backups are failing with "resource temporarily unavailable", and you have disconnected everyone from them and restarted the Time Capsule/Extreme, and the problem still exists, do the following.
1. Put the disk on a regular Mac.
2. Run Disk Utility and tell it to Repair the volume.
3. Mount the disk image to make sure it's working (just double click in the finder, let it open, the eject it). I recommend this because the first mount seemed to take forever, so I suspect it's better to do it locally than after returning it to the network--besides, you want to make sure it's working.
4. Return the disk to the Time Capsule/Extreme
5. Watch as everything works as it should.
6. Post here that it worked (or not).
No, I have no idea how to do this with a Time Capsule disk. Can you slap it on a Mac? Or do you have to pull the disk and put it in a different enclosure? Someone please comment on the correct solution in this discussion thread.
THE LONG VERSION
You try to do a backup to a remote Time Machine disk image and it fails. The error message (if you find it) looks something like this:
..."Bulimia_001451
97ba38.sparsebundle" failed with error 35. Resource temporarily unavailable
No amount of futzing remotely with Disk Utility can fix the image.
That's because the problem isn't with the image, it's actually with the disk on which the image exists. A sparseimage is actually a folder (with a lot of files in it). I realized what was going on when I did an 'ls -l' from the shell of /Volumes/xxx/yyy.sparcebundle/token and got back that same "resource temporarily unavailable" error. So this has nothing to do with the image per-se, something is wrong with the backup volume.
My backups are to disks hanging off an Airport Extreme, if you are using a Time Capsule I'm not sure how you're going to do this. I simply pulled the disk off the Extreme, put it on my laptop, ran Disk Utility and told it to repair the disk.
Oct 10 01:11:59 Shadowfax Disk Utility288: Starting repair tool:
Oct 10 01:12:00 Shadowfax Disk Utility288: Checking Journaled HFS Plus volume.
Oct 10 01:12:00 Shadowfax Disk Utility288: Checking Extents Overflow file.
Oct 10 01:12:00 Shadowfax Disk Utility288: Checking Catalog file.
Oct 10 01:12:03 Shadowfax Disk Utility288: Checking multi-linked files.
Oct 10 01:12:03 Shadowfax Disk Utility288: Checking Catalog hierarchy.
Oct 10 01:12:06 Shadowfax Disk Utility288: Checking Extended Attributes file.
Oct 10 01:12:06 Shadowfax Disk Utility288: Incorrect number of Access Control Lists^[
Oct 10 01:12:06 Shadowfax Disk Utility288: (It should be 0 instead of 22493)
Oct 10 01:12:06 Shadowfax Disk Utility288: Checking volume bitmap.
Oct 10 01:12:07 Shadowfax Disk Utility288: Checking volume information.
Oct 10 01:12:07 Shadowfax Disk Utility288: Repairing volume.
Oct 10 01:12:13 Shadowfax Disk Utility288: Rechecking volume.
Oct 10 01:12:13 Shadowfax Disk Utility288: Checking Journaled HFS Plus volume.
Oct 10 01:12:13 Shadowfax Disk Utility288: Checking Extents Overflow file.
Oct 10 01:12:13 Shadowfax Disk Utility288: Checking Catalog file.
Oct 10 01:12:14 Shadowfax Disk Utility288: Checking multi-linked files.
Oct 10 01:12:14 Shadowfax Disk Utility288: Checking Catalog hierarchy.
Oct 10 01:12:18 Shadowfax Disk Utility288: Checking Extended Attributes file.
Oct 10 01:12:18 Shadowfax Disk Utility288: Checking volume bitmap.
Oct 10 01:12:19 Shadowfax Disk Utility288: Checking volume information.
Oct 10 01:12:19 Shadowfax Disk Utility288: The volume Bulimia Backup was repaired successfully.
"Incorrect number of Access Control Lists" is the error that is causing this problem. Both the Extreme and the Time Capsule have their own versions of MacOS disk repair utilities, but they are running on NetBSD, not OSX, so presumably there are subtle differences in the code. In this case they are NOT CATCHING THE DISK ERROR. They report no problems at all with the disk (I've checked the log files), but in fact it does have a problem. One of the files inside the disk image folder has an error which prevents it from being opened, or even stat'ed, and as a result the disk image can't be mounted.
Na dit te hebben gedaan is de backup gelukt!!
In log van console kreeg ik de volgende melding;
Jan 16 09:46:31 xxx/System/Library/CoreServices/backupd[2593]: Copied 3103 files (3.7 GB) from volume Macintosh HD.
Jan 16 09:46:39 xxx/System/Library/CoreServices/backupd[2593]: Copied 3104 files (3.7 GB) from volume Backup of xxx.
Jan 16 09:46:39 xxx/System/Library/CoreServices/backupd[2593]: No pre-backup thinning needed: 319.1 MB requested (including padding), 354.38 GB available
Jan 16 09:47:41 xxx/System/Library/CoreServices/backupd[2593]: Copied 948 files (58.1 MB) from volume Macintosh HD.
Jan 16 09:47:47 xxx/System/Library/CoreServices/backupd[2593]: Copied 949 files (58.1 MB) from volume Backup of xxx.
Jan 16 09:48:00 xxx/System/Library/CoreServices/backupd[2593]: Starting post-backup thinning
Jan 16 09:48:00 xxx/System/Library/CoreServices/backupd[2593]: No post-back up thinning needed: no expired backups exist
Jan 16 09:48:00 xxx/System/Library/CoreServices/backupd[2593]: Backup completed successfully.
Jan 16 09:48:03 xxx/System/Library/CoreServices/backupd[2593]: Ejected Time Machine disk image.
[
Voor 11% gewijzigd door
Maarten Blom op 16-01-2009 09:54
]