USB Hard Drive set up as Time Machine - Unmounts randomly

Add me to the list. I started experiencing random disconnects of my TM drive about a month or so ago in 10.5.6 on my G5. The first two times it happened TM seemed to be choking on a particular directory it was trying to delete. The only way I knew of to get the drive to mount was to restart because no disk utility could see it.

After the second occurrence, I shut everything down and then turned off the TM drive's enclosure. When I restarted Time Machine was able to successfully backup, and it has been backing up for the last couple of days until tonight. This evening it disconnected during the finishing phase of the TM backup, and I got an improper disconnect warning.

The drive in question is a Samsung Spinpoint 750GB drive. I know the problem is definitely related to TM because I have an identical Spinpoint drive in the same enclosure that has never disconnected. The enclosure is a generic eSATA enclosure connected to a generic eSATA card that uses a Silicon Image driver. Strangely, OS X identifies the drives connected to this card as being SCSI drives instead of SATA drives. I had no issues with the drives until a month ago, but I don't recall making any substantial changes to my Mac then. I'd like to fix this problem. Not only is it annoying, it's partially defeating the purpose of TM.

I'm having problems with a WD External 1TB drive which is on FW800, it will randomly ummount when a USB device is connected or disconnected. WD are blaming the computer, but there are many people having the same problem on the mac with different drives?

I dont even know where to start, I cant iso test it as I dont have access to another mac.

It's most likely a problem with the Segate drive, as they're prone to un-mount due to some firmware driver thingy. Have you tried re formatting the drive?

Me too. I also get random "Device Removal" errors with an external 1TB USB Fantom drive. 😟

I guess the question is when will Apple do something to fix this. The problem did not exist using the Firewire connection but it happens constantly with the USB2, either connected through a USB port or directly into the computer. I also found that the Time Machine icon will spin constantly but nothing gets backed up. The only way I could fix this was to reboot, which I won't do every hour.

For what it's worth, I had the same problem with an IOMEGA external drive. I tried just about everything others have mentioned here. Finally, it appeared that along the USB chain connected to an USB 2.0 hub I also had an old iPod that had frozen. Once removed, everything was back to normal. No more silly unmount events. My 2 pence.

I have had it unmount randomly as well but with eSata on a Simpletech 1TB quad interface. I think it is definitely a TM error. I had the TM backup on another drive working just fine until I moved or accessed the backup from another drive. The working drive was a Mybook 500gig on FW800.

Having the same problem as above. G4 powermac using a 1TB iomega external with Time machine. When plugged in to the *USB 2.0* port I get the aforementioned disconnects, also when attempting to retrieve the data I get a permissions error and then the disconnect. I have managed to retrieve small files but nothing over a 1GB,

Tried resetting permissions, running disk repair. Diskwarrior showed many errors but could not rebuild the directory. Probably a good thing as DiskWarrior most likely would have destroyed the data.

Currently: I switched to the *USB 1.0* port, so far it working but it's slooooooooooooooow. I will update when the copy completes or fails... 8hrs from now.

At this point I'm thinking it was the last OS update. If this fails, or even if it doesn't I may try going back to 10.5.

Message was edited by: Jon Spratlin

Hi, don't know if anyone's still Reading this, but i've got a similar but slightly different problem: I have lots of stuff, including 3 external drives connected to my MacBook through a powered 13 port USB 2.0 hub. Everything works fine until the time machine backup kicks in and then a particular one of my external drives keeps unmounting. I find this quite bizarre. I have previously had a similar problem where by the drives was unmounting due to insufficient power, but there's no reason I can imagine why there'd be a greater drain on power when the time machine backup starts, unless it's simply because it suddenly tries to power up and use all 3 drives at the same time to back them up. Either that or what... Some sort of internal bandwidth problem?

Update from my last post.

Time Machine works fine on *USB 1.0* no disconnects, it is of course painfully slow.

The only other variable preventing me from pinning this on Time Machince versus USB 2.0 is that I'm using a 3rd party PCI card for my USB 2.0

When I get a chance I'll drop back to 10.5 nd see if that helps.

Happend again, twice in a row. It's a Fantom Drives Green Drive 1TB

+10/09/09 10:04:29 PM kernel disk1s3: device/channel is not attached.+
+10/09/09 10:04:29 PM kernel jnl: disk1s3: flushing fs disk buffer returned 0x6+
+10/09/09 10:04:29 PM kernel disk1s3: media is not present.+
+10/09/09 10:04:29 PM mds[21] (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V1/Stores/83F2C0EC-8D6F-4D93-9E5C-FA6E1AB77E57)(E rror) IndexCI in store stream_flush:store_streamflush write err:22+
+10/09/09 10:04:29 PM kernel jnl: disk1s3: do jnlio: strategy err 0x6+
+10/09/09 10:04:29 PM mds[21] (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V1/Stores/83F2C0EC-8D6F-4D93-9E5C-FA6E1AB77E57)(E rror) IndexCI in _ContentIndexSyncIndexBulk:Sync error live.1.+
+10/09/09 10:04:29 PM kernel jnl: disk1s3: write journalheader: error writing the journal header!+
+10/09/09 10:04:29 PM kernel disk1s3: media is not present.+
+10/09/09 10:04:29 PM mds[21] (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V1/Stores/83F2C0EC-8D6F-4D93-9E5C-FA6E1AB77E57)(E rror) IndexGeneral in notify_lowspace:low space for device 234881029 (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V1/Stores/83F2C0EC-8D6F-4D93-9E5C-FA6E1AB77E57)+
+10/09/09 10:04:29 PM kernel ystem.fs] [ErrType IO] [ErrNo 6] [IOType Write] [PBlkNum 64613456] [LBlkNum 0] [FSLogMsgID 865659127] [FSLogMsgOrder First]+
+10/09/09 10:04:29 PM mds[21] (Error) Volume: LOW DISK SPACE device:234881029+
+10/09/09 10:04:29 PM mds[21] (/Volumes/Custodian/.Spotlight-V100/Store-V1/Stores/0E5C6A6C-4ECB-48AB-A25C-57A 53F507DBF)(Error) IndexStore in SIStoreDirytySDBChunks:Error storing dirty sdb pages: 22+
+10/09/09 10:04:29 PM kernel disk1s5: media is not present.+
+10/09/09 10:04:29 PM mds[21] (/Volumes/Custodian/.Spotlight-V100/Store-V1/Stores/0E5C6A6C-4ECB-48AB-A25C-57A 53F507DBF)(Error) IndexCI in CIMetaInfoSync:sync err: 22+
+10/09/09 10:04:29 PM kernel disk1s5: media is not present.+
+10/09/09 10:04:29 PM kernel disk1s5: media is not present.+
+10/09/09 10:04:29 PM kernel disk1s3: media is not present.+
+10/09/09 10:04:29 PM kernel disk1s3: media is not present.+
+10/09/09 10:04:29 PM kernel jnl: disk1s5: close: journal 0x6f92d10, is invalid. aborting outstanding transactions+
+10/09/09 10:04:29 PM kernel jnl: disk1s5: close: journal 0x6f92d10, is invalid. aborting outstanding transactions+
+10/09/09 10:04:30 PM kernel jnl: disk1s3: close: journal 0x6f92e08, is invalid. aborting outstanding transactions+
+10/09/09 10:04:30 PM kernel jnl: disk1s3: close: journal 0x6f92e08, is invalid. aborting outstanding transactions+
+10/09/09 10:04:38 PM kernel jnl: disk1s3: replay_journal: from: 28684800 to: 32583168 (joffset 0xa5f000)+
+10/09/09 10:04:40 PM kernel jnl: disk1s3: journal replay done.+
+10/09/09 10:04:40 PM kernel jnl: disk1s3: journal replay done.+
+10/09/09 10:04:40 PM kernel HFS: Removed 2 orphaned unlinked files or directories+
+10/09/09 10:04:40 PM fseventsd[27] event logs in /Volumes/Time Machine Backups/.fseventsd out of sync with volume. destroying old logs. (1153 3 1153)+
+10/09/09 10:04:40 PM kernel jnl: disk1s5: replay_journal: from: 3799040 to: 4537344 (joffset 0x159d000)+
+10/09/09 10:04:40 PM fseventsd[27] log dir: /Volumes/Time Machine Backups/.fseventsd getting new uuid: DEDBAB13-E906-421A-A287-62E751978F38+
+10/09/09 10:04:41 PM kernel jnl: disk1s5: journal replay done.+
+10/09/09 10:04:41 PM fseventsd[27] event logs in /Volumes/Custodian/.fseventsd out of sync with volume. destroying old logs. (419 1 1154)+
+10/09/09 10:04:41 PM fseventsd[27] log dir: /Volumes/Custodian/.fseventsd getting new uuid: ED0E5A85-067E-4DA5-B64F-2A963C00028E+
+10/09/09 10:06:57 PM kernel disk1s3: media is not present.+
+10/09/09 10:06:57 PM kernel disk1s3: media is not present.+
+10/09/09 10:06:57 PM kernel disk1s3: media is not present.+
+10/09/09 10:06:57 PM kernel e Write] [PBlkNum 3501328] [LBlkNum 0] [FSLogMsgID 827976120] [FSLogMsgOrder First]+
+10/09/09 10:06:57 PM mds[21] (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V1/Stores/83F2C0EC-8D6F-4D93-9E5C-FA6E1AB77E57)(E rror) IndexStore in SIStoreDirytySDBChunks:Error storing dirty sdb pages: 22+
+10/09/09 10:06:57 PM mds[21] (/Volumes/Time Machine Backups/.Spotlight-V100/Store-V1/Stores/83F2C0EC-8D6F-4D93-9E5C-FA6E1AB77E57)(E rror) IndexCI in CIMetaInfoSync:sync err: 22+
+10/09/09 10:06:57 PM kernel disk1s5: media is not present.+
+10/09/09 10:06:57 PM kernel disk1s5: media is not present.+
+10/09/09 10:06:57 PM kernel disk1s5: media is not present.+
+10/09/09 10:06:57 PM mds[21] (/Volumes/Custodian/.Spotlight-V100/Store-V1/Stores/0E5C6A6C-4ECB-48AB-A25C-57A 53F507DBF)(Error) IndexStore in SIStoreDirytySDBChunks:Error storing dirty sdb pages: 22+
+10/09/09 10:06:57 PM mds[21] (/Volumes/Custodian/.Spotlight-V100/Store-V1/Stores/0E5C6A6C-4ECB-48AB-A25C-57A 53F507DBF)(Error) IndexCI in CIMetaInfoSync:sync err: 22+
+10/09/09 10:06:57 PM kernel jnl: disk1s3: close: journal 0x6f92e08, is invalid. aborting outstanding transactions+
+10/09/09 10:06:57 PM kernel jnl: disk1s3: close: journal 0x6f92e08, is invalid. aborting outstanding transactions+
+10/09/09 10:06:58 PM kernel jnl: disk1s5: close: journal 0x6f92d10, is invalid. aborting outstanding transactions+
+10/09/09 10:07:00 PM kernel jnl: disk1s3: replay_journal: from: 32583168 to: 33081344 (joffset 0xa5f000)+
+10/09/09 10:07:00 PM kernel jnl: disk1s3: journal replay done.+
+10/09/09 10:07:00 PM kernel jnl: disk1s5: replay_journal: from: 8038912 to: 8768512 (joffset 0x159d000)+

Upgraded to Snow Leopard 10.6 - Problem still occurs.

I changed the cable connecting the USB drive to the computer. The problem went away.

Sorry about the false alarm.