Disk Drive ejecting itself

Hi all - I think this issue is far more down in the depths than simple preferences.

After giving up on my (now utterly useless) mac mini server I wired up my time machine drives to my Time Capsule instead. After 2 days happilly getting all my backups in place (I have 2 backup drives running and 3 machines to back up) and being about 80% there the Time Capsule decide to eject the external drives too.

This is just too much. Time capsule is a great idea and, to be honest, is one of the few things that actually keeps me using a mac at all. After all the hassle, cost in lost hardware and lost work hours (I estimate this issue has cost me over £1000 in time and parts costs so far) I am seriously considering moving back to a Linux / Windows hybrid system.

Your backups need to be reliable but with this I have no faith in my backups or even the safety of my own data!

As a follow up to my post of April 25, 2014. I have, in the last three weeks, switched out the one drive that was "Green" for a "NAS" drive (both Seagate). Since swapping out and transferring data between the drives (fyi the "green" drive is no longer attached via eSATA, Thunderbolt or USB to my iMac), I have not experienced a single drop or ejected drive.

I can say, right now, that replacing the Green drives with proper NAS drives seems to have solved my issue.

Cheers

I'm joining in with my observations, more for moral support than anything else.

I've got a mid 2012, 13-inch MBP. I replaced the optical drive with a 1Tb HGST Travelstar. I used an MCE Technologies Optibay kit. I wanted to carry an on board mirror of the boot drive.

The computer treats the new drive as a mounted, external drive. So, it suffers all the same issues and symptoms described above.

I can however add the following observations: Seems to be large file transfers causing the issue. A file here and there rarely (never?) causes the drive to eject. A minute before the drives ejected, a click is heard from the upper right hand side, near the power switch. A moment later, the dreaded ejection occurs.

The ejection issue is independent of the drive format. I tried formatting the drive as NTFS, and use Paragon. Slllooowww, is the only way to describe the result and still the drive unmounted. By the way, Carbon Copy Cloner wrote to the drive formatted this way.

So, I thought for a while it was caused by my mac choking on a hidden Windows file type. I moved some of the offending folders to a windows machine and unhid the files. Nothing. BTW, I maintain primarily video files, all formats.

Just thought I'd add my experieince, it is very frustrating. It makes externally backing up the machine impossible and untrustworthy. In this day and age of the cloud experience that Apple so much wants us to embrace, how can this be unaddressed? My Windows machines perform flawlessly in this regard. I no longer see an advantage in owning the pricier Apple equipment.

I guess I'm hoping for an answer to this of some sort. Rumors that Apple will be addressing this in the future.... The cable and drive suggestions wont work for me, its internal now, connected directly to the mobo.

I'm considering replacing OSX with Windows. Thoughts there? How about the second drive as the Windows drive?

Thanks for listening,

J

An interesting post, since this is one of the first that shows this problem persists even when using internal hardware connections only. A Mac owner wanting to install Windows - gasp!!! A pretty pass, that! Just sell it and buy a real PC where your Windows power management drivers (one of the biggest banes of Windows) have a snow ball's chance of working. If you use your Mac for business, run - don't walk - to change your hardware.

FYI: This thread was created for OSX Snow Leopard, which Apple has quietly, with no formal announcement, ceased to support with security updates as of last September. So no fix for the drive issue is forthcoming for the original post - there never was. Apple has a long-standing habit of forcing customers to update their OS after only four years (if they care about security) and a history of orphaned hardware (in terms of network connectivity) far inferior to Windows. XP had an 11 year run before Microsoft pulled the plug; it offered an architecture a business could rely on for a decade. In 2003, we were still running OSX 10.2 on G4s, with OS 9 users (and all their software) freshly jettisoned to the dust-bin or forced to limp on with "Classic."

So here we have this dismal support for a very basic hardware problem, which now has outlived the OS where this thread was started. Given this track record, coupled with higher prices and short architecture lifespans, is it any wonder that Apple still sells fewer computers than its PC rivals? I can't say I am hopelessly in love with this Lenovo or Windows 8, but my backups run.

For those that want to press on or have already invested in Mac hardware and software, read my earlier posts, echoed by dcsang and others, and get a NAS. You can then use ethernet to talk to your backup repository. A NAS running the current SMB2 protocols generally does not care what current flavor you are running (Mac 10.7+, Win Vista-7-8, Ubuntu 13.x+). With OSX 10.9, Apple has, once again, quietly and with no formal announcement, abandoned a core technology: it is ditching AFP file sharing for the current Windows and Linux standard: SMB2. Note that your AFP file shares will be (you guessed it!) in the dustbin, as is Snow Leopard, which, like XP, can't do SMB2 or DFS shares. So you might think that in the near future, the NAS or Win/Linux file server approach for storage looks even better than before. But apparently not all is happy in Maverick-land:

http://www.zdnet.com/mavericks-smb2-problem-and-fixes-7000022519/

It looks like Maverick has some trouble with SMB2. But for a home user with no heavy network security requirements, one can still dial this back to SMB1, which also means your older 10.6 and earlier Mac equipment will still work with it.

So maybe it really is time to forget Thunderbolt, Firewire, and USB and throw those cables in the shoe box to join the old SCSI junk. End of story.

zI appreciate the response

There are mountains of these types out there (e.g, google "optibay, macbook pro, eject unexpectedly) dating back the last few years. I found one guy with the identical problem, including the make and model of the drive he installed.

I mirror your response. I have been able to get the mac to back to my Synology DS1513+ using Carbon Copy Cloner. That seems to work pretty solid even though Tme Machine failed me there.

And finally, the optibay guy actually wrote me back, at least feigning interst in my pitiful problem.

Ths was an attempt to go all in with Apple an iPad, iPhone and then the MacBook Pro. I have been let down.

Work has provided me with a Galaxy S4. I'm able to do side by side comparisons....

My girlfriend has a Kindle HDX. We found an issue on it preventing the email from switching between plain and HTML. Over te Kindle speaker the tech explained there was no way to do that, it was an omission in his opnion and he would elevate this for an update revision. Again, perhaps another veiled attempt to passivate the masses, but you feel better after an expereience like that.

I'm going to get out of the exclusive Apple business.

If this is an energy management issue, then switching the HDD to an SSD might remedy it.

Or maybe not....

Any thoughts there?

Well everyone, after 62 pages, 10 months and multiple itterations of the OS, I am officially tapping out.

I just got back from outfitting my studio with six Dell workstations. 😟 (three to follow shortly)

I cant wait any longer to upgrade my workstations, and my people in the field cant rely on NAS or cloud solutions, and after Genious Bar trips, phone support and online searching, I can see that there is no fix coming for the people experiencing this. I dont know if not enough of us are seeing this for Apple to step in, but in any event, I cannot run a business this way.

I cant have questions as to if someone could or could not backup their system while working away from the office, or have client files stuck on a suddenly inaccessible external harddrive.

I cant constantly be emailing out workarounds to people to try and figure out how to make something that should just work do what it should.

I know that Apple will never see this, but I just feel like I needed to say it. All phone/tablet upgrades are going into Android devices and tablets. Trying to run an Apple shop was fine, but I dont feel like dealing with the inevitable issues that ocurr with a mixed Apple/Windows environment. Especially not when doing so would be only to show loyalty to a company that I feel has forgotten me (all of us).

Im not even angry. Im sad.

My thunderbolt docking solution has been 'eject free' for about 1/2 year so I forgot about the problem.

Is there still a eject problem with Mavericks?

The answer:

I took the 1TB HGST Travelstar 7K1000 and gave it a firmware downgrade. HGST emailed me application to set the port to either, Sata I, II, or III. Reducing it to Sata II solves the problem.

No more dropping.

Has somethgn to do with Apple/intels first venture into the Sata III world back then. I used the less expensive Opti bay kit. Not the more expensive Extreme they have created to solve the problem.

Works great and is plenty fast for a seconday drive.

62 pages and the issue is still there. Apple is one of the few companies that leave bugs on their software for years. Incredible. I'll avoid connecting this new HDD to this computer again (and use my Windows computer instead) because I've lost all info on two other drives because of this bug,.

Latest update (10.9.4) lists "Improves the reliability of waking from sleep"

... anyone know if this has resolved the issue with external Thunderbolt drives?

About the OS X Mavericks v10.9.4 Update

Hi!

I guess it's my time to write now. Don't know if it's smart to wake up this thread again, but here I go.. Btw, my native language is not English, therefore excuse my rusty English :-)

I have a 4-bay IB-RD3640SU3 from ICYBOX with 4x Seagate 2TB 7200rpm Barracuda disks in raid-5, and I connect it to my brand new iMac with USB-3. I never had any disconnects in Windows using eSATA->eSATA, not even once, for the record..

Just before I got my iMac I tried USB-3 in Windows to see if it's any faster than eSATA (it wasn't), but I discovered it sometimes disconnected while copying large amounts of data from 4-bay to another disk. I never used power saving features on Windoze but I did get some random disconnects now and then (5-10 minutes in standby) after trying USB-3. In my case the IB-RD3640SU3 spins down disks after a period of inactivity, even tho power saving is disabled, regardless eSATA or USB-3. I really don't know if this problem is because of USB-3 and cheap/poor controller inside 4-bay, cable or OS. I changed back to eSATA to back up my 4-bay before formatting for OS X and had no problems.

As mentioned, I just got my brand new iMac, and the first thing I did was to connect my 4-bay raid-5. For the record, I formatted drive with GPT (MAC Journ (not case, no crypt). When I copied large amounts of data I had the same issue in OS X as well, esp adding about 2000 jpegs to iPhoto. When the drive was in standby it also disconnected after a while, sometimes just before "sleep" (1hr inactivity) and sometimes after 5 minutes after last usage.. And...sometimes I get the "disconnected" warning just after i "wake up" the iMac. WTH..!

I believed that my USB-3 cable was ruined, and so I change the cable. New cable is now 1.5 meters vs "old" one which was 3.0 meters, and now I notice a decrease in disconnects. Don't get me wrong, its still not usable in my opinion!

I also get disconnects under huge workloads, like thousands of smaller files being copied/read from 4-bay. I haven't tried with larger files since I mostly edit photos. Same story with disconnects after inactivity, but its a lot improved vs when I used a longer cable.

OMG, soo many variables! :-(

Updating firmware doesn't work in my case, my disks is being recognized as 1 large (almost 7 TB) volume by OS (reason; Raid-5 GPT), and changing from Raid-5 to JBOD will ruin the parity of Raid-5. I doubt upgrading firmware is going to make any diff since reverting to eSATA in Win fixed the problem. I'm kinda' mental stuck on having 4-bay with a terrible USB-3 controller, or worse, OS X having problems with the whole ICYBOX "want-to-spin-down-when-idle .. while having-pms-with-USB-3 and/or USB-3 controller".

Maybe Keep Drive Alive can help my random disconnects, but I doubt it would help against the hibernation sleep/awake disconnects. I'm thinking about running Keep Drive Alive and Jettison, but I dislike to install any 3rd party software which may cause other problems to my workstation. Well, esp 2 potential softwares like that in one go. So, are there any people inside here running those 2 applications side-by-side? Do you run in to any issues performance wise?

The thread has been "sleeping" a long time now, did Apple provide any patch that I didn't get? :-S

Is there any people using Thunderbolt->eSATA adapters in here? Which one is recommended to use, if any? I've heard a lot of crp about Lacie, and Kanex is not a brand I know anything about. I believe if I use eSATA like I did earlier in Windows my problems will go away, most likely.

FYI, right before the 4-bay takes a dive under workload I also get the spinning beach ball of death.

I have done some research about this problem and I don't agree that Apple "sleeps in their class". They released 3 updates regarding this issue in 1 year, and you also have to think about how many brands there is, how many different stories/problems, how many different setups, how many different.. Etc. What I dont understand is that my WD Elements 500GB drive I'm using for TimeMachine have _no_ problems like this, and this drive use power from USB-3, while my 4-bay use external power! :@ WTH ?!?!

I just experienced this on a brand new 1 TB G-drive connected on the Thunderbolt cable it came with, fresh out of the box 5 minutes ago and about 10 minutes into a Carbon Copy Cloner session. Attached to a 13" MacBook Air, 8Gb 1600 Mhz DDR3, 1.7 Ghz Intel Core i7, OS Version 10.9.4. I'm cloning the built-in 256 Gb SSD. On my screen CC Cloner reports "CCC aborted the task because the destination volume disappeared". I have notifications for each of the 3 partitions on the G-drive that were simultaneously ejected. No physical contact with the drive, cable, or laptop.

My power settings are set to never put the disk to sleep. I've seen the auto-eject problem now twice in the span of less than 10 minutes.

I am able to easily reproduce the auto-eject problem during Carbon Copy Cloner operations even while booted up in Safe Mode. However, I did run the procedure you suggested, do you see any kernel extensions in the diffs I captured?

# used AWK to strip out all but column 6 in the original files

diff full.txt safe.txt

25a26

> com.apple.driver.AppleIntelLpssDmac

29d29

< com.apple.driver.AppleIntelLpssDmac

36d35

< com.apple.driver.AppleHSSPISupport

38a38

> com.apple.driver.AppleHSSPISupport

45a46

> com.apple.driver.XsanFilter

48d48

< com.apple.driver.XsanFilter

51a52

> com.apple.driver.AppleFileSystemDriver

63a65,66

> com.apple.iokit.SCSITaskUserClient

> com.apple.iokit.IOUSBHIDDriver

73d75

< com.apple.driver.AppleCameraInterface

75,77c77

< com.apple.driver.AppleSMCLMU

< com.apple.iokit.IOHDAFamily

< com.apple.driver.AppleHDAController

---

> com.apple.driver.AppleSMBusPCI

83,91c83,86

< com.apple.driver.AppleThunderboltIP

< com.apple.driver.AppleHWAccess

< com.apple.kext.OSvKernDSPLib

< com.apple.iokit.IOAudioFamily

< com.apple.vecLib.kext

< com.apple.driver.DspFuncLib

< com.apple.driver.AppleHDA

< com.apple.iokit.IOSurface

< com.apple.driver.AppleIntelHD5000Graphics

---

> com.apple.driver.IOPlatformPluginLegacy

> com.apple.driver.ACPI_SMC_PlatformPlugin

> com.apple.driver.AppleGraphicsControl

> com.apple.driver.AppleMuxControl

92a88

> com.apple.driver.AppleGraphicsDevicePolicy

96,103c92

< com.apple.driver.ApplePlatformEnabler

< com.apple.iokit.IOUserEthernet

< com.apple.driver.AppleUpstreamUserClient

< com.apple.iokit.IOSerialFamily

< com.apple.iokit.IOBluetoothSerialManager

< com.symantec.kext.ndcengine

< com.symantec.kext.ips

< com.apple.driver.AudioAUUC

---

> com.apple.iokit.IOSurface

106,110d94

< org.virtualbox.kext.VBoxDrv

< org.virtualbox.kext.VBoxUSB

< org.virtualbox.kext.VBoxNetFlt

< org.virtualbox.kext.VBoxNetAdp

< com.symantec.kext.SymAPComm

Lest this lovely little thread die out, it looks like I just enountered the problem while running a CarbonCopyCloner backup from a MacBook Pro Retina mid-2012 (Yosemite OS X 10.10 14A388a) to a 4 TB WesternDigital MyBook via gigabit wired ethernet all the way. FWIW, two large Parallels VM files backed up okay (~53 and 36 GB,) but the job terminated after 8 1/2 hours while backing up with: "DISKCONT| 10/27 14:52:02 [INFO] Volume unmounted: "CarbonCopyCloner" (//Hal@WDMyCloud._afpovertcp._tcp.local/CarbonCopyCloner)"

noted in the ccc_debug.log file. It looks like this is still happening...I just bought CCC and am new to this issue.