When I set up a user share and choose "time machine" under the AFP protocol and then turn off the other CIFS sharing, NFS, etc. and then tell time machine to use that volume, it will start backing up for a few minutes and then my Mac crashes.  Any ideas or logs I can provide?  Thank you.

Use Disk Utility on the Mac to repair the Mac disk and repair permissions.

Use Disk Utility on the Mac to repair the Mac disk and repair permissions.

 

I did have some permissions issues and had to also repair the disk.  I am backing up 3.14GB of 555.1GB (about 16 hours remaining).

 

We'll see how it goes.  I am not out of the woods yet as I have had it going like this a couple of times and then it just stopped because the volume "goes offline" or something like that.  I think it is the one called "Time Machine Backups" but my user created "tm" volume is still mounted.  I will let everyone know the progress.

It failed.  "Time machine couldn't complete the backup to "server".  An error occurred while copying files.  The problem may be temporary.  If the problem persists, use Disk Utility to repair your backup disk.  The "tm" volume is still mounted but the "Time Machine Backups" volume that it mounts when doing the backup (green one as opposed to the regular blue one) is unmounted.  Advice?

 

 

Start a new one. It is reliable once its working.

I have tried that - 3 times now.  Same result.

Maybe I have something configured wrong on the unRAID server?

Post a syslog after the Mac has crashed.

The Mac is not crashing now.  It's just that the volume that Time Machine mounts called "Time Machine Backups" is not there.  It gets unmounted for some reason.

The Mac is not crashing now.  It's just that the volume that Time Machine mounts called "Time Machine Backups" is not there.  It gets unmounted for some reason.

 

This one:

 

Sep 19 19:26:18 Davids-iMac-2.local com.apple.backupd[982] <Notice>: Disk image /Volumes/tm-1/David’s iMac (2).sparsebundle mounted at: /Volumes/Time Machine Backups

 

It's the sparsebundle image that gets created on the user defined "tm" volume which is located on the unRAID server.

 

 

I know this is a silly question, but I have to ask.

 

Is the time machine share on unRAID a single drive or is it set to use more then one disk?

I think you get where I am going with this.

 

Not a silly question at all.  It is spanned on multiple disks which I THOUGHT was acceptable so you could have one huge disk.  So is that the culprit?

 

I am seeing this in the logs:

 

<Debug>: hfs: Runtime corruption detected on Time Machine Backups, fsck will be forced on next mount.

com.apple.backupd[1173] <Error>: Stopping backup.

com.apple.backupd[1173] <Error>: Error: (-8062) SrcErr:NO Copying /bin/[ to (null)

com.apple.backupd[1173] <Notice>: Copied 279333 files (42.2 GB) from volume Macintosh HD.

com.apple.backupd[1173] <Error>: Copy stage failed with error:11

com.apple.backupd[1173] <Error>: Backup failed with error: 11

com.apple.kextd[12] <Error>: /Volumes/Time Machine Backups: error opening /usr/standalone/bootcaches.plist: Input/output error.

mds[38] <Alert>: (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V2/BF94D866-FCEF-4A41-B428-25E5EB5CFB48)(Error) IndexCI in _Bool store_stream_init(store_stream_t *, fd_ptr, _Bool):store_stream_init err:2

mds[38] <Alert>: (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V2/BF94D866-FCEF-4A41-B428-25E5EB5CFB48)(Error) IndexCI in int indexPrepareForSyncBulk(uint32_t, ContentIndexPtr *, uint8_t *, int *, int, _Bool, _Bool (*)(void *), void *):indexPrepareForSync error: 2, live.0.

mds[38] <Alert>: (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V2/BF94D866-FCEF-4A41-B428-25E5EB5CFB48)(Error) IndexCI in int _CISyncContextSync(CISyncContextPtr, fd_ptr *):indexPrepareForSyncBulk error:2

fseventsd[48] <Error>: disk logger: failed to open output file /Volumes/Time Machine Backups/.fseventsd/636573aad860711c (Input/output error). mount point /Volumes/Time Machine Backups/.fseventsd

com.apple.kextd[12] <Error>: /Volumes/Time Machine Backups: error opening /usr/standalone/bootcaches.plist: Input/output error.

fseventsd[48] <Error>: disk logger: failed to open output file /Volumes/Time Machine Backups/.fseventsd/636573aad860711c (Input/output error). mount point /Volumes/Time Machine Backups/.fseventsd

com.apple.backupd[1173] <Notice>: Ejected Time Machine disk image: /Volumes/tm-1/machine.sparsebundle

kernel[0] <Debug>: AFP_VFS afpfs_unmount: /Volumes/tm-1, flags 0, pid 1469

kernel[0] <Debug>: AFP_VFS afpfs_unmount : We are the last mnt/sbmnt using volume /Volumes/tm-1 0xffffff81ad288008

kernel[0] <Debug>: ASP_TCP do_thread_read: no reqInfo found for reqID 1

kernel[0] <Debug>: AFP_VFS afpfs_unmount : We are the last volume using socket /Volumes/tm-1 0xffffff81ad288008

com.apple.backupd[1173] <Notice>: Ejected Time Machine network volume.

kernel[0] <Debug>: AFP_VFS afpfs_unmount : afpfs_DoReconnect sent signal for unmount to proceed

 

 

It should not matter since a TM sparce-bundle is just a directory tree of files as far as unRAID is concerned. But it may be an issue. Additionally, TM will grow the backup until all of the space is filled and with multiple backups competing for space there is no way to control how much space each client gets. I suggest that you assign each TM client an independent share, each size limited appropriately. Limit each TM share to a single disk.

Yeah, not the answer I was hoping for.  I may just throw a 3TB disk into my Time Capsule instead of the 1TB it has in there and go from there.