Dead Mac! S.M.A.R.T. Status: Failing. Can't boot!

[Chunk1978]

today my 27" iMac i bought a year and a half ago suddenly because really sluggish.

i checked the processes to see what was causing the problem. nothing seemed to be taking up large amounts of resources so i rebooted. again, sluggish so i rebooted again. soon thereafter i decided i should try and repair disk permissions so i fired up Disk Utility and saw the HD listed in red! "S.M.A.R.T. Status: Failing" and something about taking the computer in for repairs as soon as possible.

before i lug this beast downtown to the Apple Store, i decided to do a bit of testing, research, but nothing i found on google seemed to exemplify what i was experiencing. sure, there are lots of posts about failing S.M.A.R.T. drives, but there is more:

now the computer doesn't boot up. i get the white background with gray apple logo and the spinning gear, but it will spin forever and not actually boot.

i can't bootup with any DVD. i've tried the install DVD that came with the computer as well as a TechTool Pro boot DVD. while pressing "C" should boot the computer from the DVD drive, the DVDs try loading and then simply eject. more, pressing "option" to choose which drive to boot from only returns one option: "Macintosh HD".

i've unsuccessfully tried to boot in Safe Mode by pressing "Shift". in doing so, the gray apple logo appears and the spinning gear. minutes later the machine will restart, only this time i will see a blinking folder icon with a question mark.

additionally, i've unsuccessfully tried to boot in Single-User Mode by pressing "Command + S" as well as Verbose-Mode by pressing "Command + V". both modes display command line prompt while booting up, but it eventually hangs and i'm never able to execute any commands. Single-User mode makes it to:

waiting for window server before finishing bluetooth setup

while Verbose Mode stalls at:

AppleAHCIDiskQueueManager::setPowerState(0x8d9b700, 2 -> 1) times out after 100103 ms

Finally, i've reseated the RAM with no luck and the Apple Hardware Test was run by starting my machine and holding "D". i've run both a regular (~10 minutes) and extended (~1 hour) testing with AHT. both tests return no hardware issues found. (??!!)

does anyone have any insight on this? if it's simply the HD that has failed, then why am i not able to boot up using a DVD and why hasn't AHT found any problems with the hardware? could it be the logic board? power supply? bad RAM? if so, again, shouldn't the Apple Hardware Test spotted these issues instead of passing everything with flying colors?

 

[Makosuke]

The original symptoms you're describing sound exactly like a failing hard drive. The failure to boot into single user mode (or any other mode from the hard drive) are, also--it's getting partway through the boot process before hanging on either a corrupt file due to disk malfunction, or the disk hardware itself failing to respond while it tries error recovery. I'll bet that if you listen very closely, you'll hear a repetitive click-click-click sound from the hard drive as it runs its internal error routines.

As for why it won't boot from an optical disc, I can't say for sure, but it's probably the same thing--the badly malfunctioning hard drive may be screwing something up at a low level, preventing the computer from booting at all. Could theoretically be a different problem, but given the original symptoms, not likely--badly malfunctioning hardware can cause just about anything to happen. Heck, I've seen failed keyboards hang the computer at power-on on a white screen.

Same goes for the AHT routine--it's mostly looking at motherboard hardware and RAM, and is notorious for not registering failing components (it's not all that thorough) so it not showing any failure tells you little if anything. The fact that you can boot into AHT--which initializes almost nothing, unlike the OS install, which is a mostly-complete OSX install and capable of hanging on bad hardware that AHT will ignore--shows that your optical disc drive is fine, and the motherboard probably is ok, too.

Bottom line, though, is that whatever is wrong, it's serious and almost certainly hardware, so you're going to need to take it in for service regardless. Assuming you have AppleCare, it's under warranty, so figuring out what exactly is wrong is their job, not yours, and they'll fix whatever it is. If it's not under warranty, that just means it'll cost a lot more if it's something other than the hard drive, but you won't know until they put it on the bench.

If it were out of warranty and you're handy with the guts of an iMac (they're not easy to work on), you could crack it open and try replacing it as a diagnostic (that's what I'd do), but that's pointless if it's under warranty, and foolhardy if you've never worked inside one before.

I hope you have a recent backup--it's pretty much guaranteed that whatever was on the drive is toast now.

 

[Macman45]

I Salutary Tale

For running disk utilities at least once a month, and especially if you notice a decrease in performance.

Pretty much everything is covered above, and unless you are very handy, I'd leave it to Apple.

If you have critical data that isn't backed up, you could go down the third party data recovery route, but it costs $$$$$$ so I hope you have a recent TM backup for the old drive....As described above, it's bricked now.

 

[Chunk1978]

i'm not covered by Apple Care, so instead of going to apple and having them charge me a million dollars to install a new HD, i found someone who came and did it for me for much cheaper. that being said, i'm not sure that was the right approach since he wasn't exactly what i would call "gentle" with my hardware.

the super drive reads CDs and some DVDs, but not all. i'm not sure if it needs to just be cleaned or if it's broken.

the computer now turns on at least, but there are still and handful of noticeable problems.

1. when the screen goes to sleep it doesn't actually turn off, but animation freezes and the mouse cursor disappears. moving the mouse to wake the display from sleep resumes animation (such as a progress bar during a software install) and the mouse cursor reappears.

2. moving windows around and other animations such as screen savers are very choppy.

3. launching Photo Booth displays "The graphics card installed in this computer does not support Photo Booth". but iSight camera shows up in the System Profiler, as does the graphics card.

there are probably a few more things wrong with it too, but these are the ones i found so far.

any ideas? it's likely that i'll have to take it to Apple in the end but i would like to hear some thoughts on what the problems could be.

 

[Mattie Num Nums]

Failing SMART is catastrophic. I wouldn't trust anything the computer is telling you when booted to the hard drive. What I would do is try to back up whatever you can ASAP and replace the HDD.

 

[Chunk1978]

Failing SMART is catastrophic. I wouldn't trust anything the computer is telling you when booted to the hard drive. What I would do is try to back up whatever you can ASAP and replace the HDD.

the HDD was replaced

there are still problems

 

[Mattie Num Nums]

the HDD was replaced

there are still problems

Do you have an log files we can look at to try and narrow down whats going on? Check Console.app in the Utilities folder the next time you have a crash. It is time stamped. This will really help narrow down where the issue is originating.

 

[Makosuke]

the HDD was replaced

there are still problems

Where did the OS install on the replaced hard drive come from? Is it from a previous backup, or a fresh install?

If it's not a totally fresh install, I'd break out your original OS discs, reformat the drive completely, install a clean OS, and test it before migrating any of your data. If it's still acting up at that point, the problem is definitely hardware--either things were broken before and the hard drive was only a symptom, or the guy who replaced the hard drive for you screwed it up while he was in there.

If it's fine, you have corruption somewhere in the data; you can try Migration Assistant and see if it's ok this time, or move things over one at a time.

Edit: After thinking about your new symptoms a little more, it sounds to me an awful lot like the computer isn't properly communicating with the monitor.

The choppy animation when moving windows in particular is what happens when the GPU accelerated interface elements of the OS don't load, which will happen if the computer is booted in "headless" mode (no monitor attached) under 10.6, even if you later connect a monitor. Ran into that problem with my home server if I booted it without the TV on. The "graphics card does not support" message is the same; iTunes will refuse to use Coverflow, and other stuff like that.

Now, it could just be that the graphics card is fried, or that there's something wrong with the OS install so the graphics card's drivers aren't loading, but again, if you boot a 10.6 (or, I think, 10.5) system with no monitor attached, you will see exactly the same thing happen when you later attach one.

Since the monitor is also refusing to properly sleep, it could be something as simple as a loose cable internally that the guy who worked on it messed up.

Reinstalling the OS fresh is a first step to test, though.

 

[Chunk1978]

Where did the OS install on the replaced hard drive come from? Is it from a previous backup, or a fresh install?

If it's not a totally fresh install, I'd break out your original OS discs, reformat the drive completely, install a clean OS, and test it before migrating any of your data. If it's still acting up at that point, the problem is definitely hardware--either things were broken before and the hard drive was only a symptom, or the guy who replaced the hard drive for you screwed it up while he was in there.

If it's fine, you have corruption somewhere in the data; you can try Migration Assistant and see if it's ok this time, or move things over one at a time.

Edit: After thinking about your new symptoms a little more, it sounds to me an awful lot like the computer isn't properly communicating with the monitor.

The choppy animation when moving windows in particular is what happens when the GPU accelerated interface elements of the OS don't load, which will happen if the computer is booted in "headless" mode (no monitor attached) under 10.6, even if you later connect a monitor. Ran into that problem with my home server if I booted it without the TV on. The "graphics card does not support" message is the same; iTunes will refuse to use Coverflow, and other stuff like that.

Now, it could just be that the graphics card is fried, or that there's something wrong with the OS install so the graphics card's drivers aren't loading, but again, if you boot a 10.6 (or, I think, 10.5) system with no monitor attached, you will see exactly the same thing happen when you later attach one.

Since the monitor is also refusing to properly sleep, it could be something as simple as a loose cable internally that the guy who worked on it messed up.

Reinstalling the OS fresh is a first step to test, though.

hey Makosuke, thanks for taking the time to respond to my questions.

the loose cable theory sure makes a lot of sense. this guy was a pure amateur. when he replaced the HD he didn't even disconnect the monitor entirely and it was leaning out and i told him i was concerned with the cables that were still attached since they looked like they were bending.

to give you another example about how dumb this guy is, when he want to reinstall the OS onto the new HDD, i passed him my 10.6.2 install DVD, but he exclaimed in his thick Pakistani accent "i give you good deal, i install newer OS". but he ended up installing Mac OS 10.5.8! what an idiot!

i suppose it wouldn't be so bad, but since my iMac's super drive isn't able to read my Snow Leopard DVD, i've been spending time ripping it to a USB drive hoping it will actually install.

note to self: goto Apple next time.

 

[Chunk1978]

looking positive so far. i'm currently booting from Snow Leopard on from a 16GB USB Flash drive and zeroing out all data. from the boot the Disk Utility window seems to be moving fine, but the best part is that the screen actually dims and then goes to sleep properly.

crossed fingers that this is all simply a bad install of 10.5.8 and that a clean install of SL will make this beauty run like new again.

 

[Makosuke]

this guy was a pure amateur.

That is BAD news when working on a newer iMac. You may have seen this yourself if you were watching him work, but it is VERY challenging to get into them; even having a fair amount of experience dismantling laptops, and with an iFixit illustrated takeapart guide in front of me, I have to be very careful, and it's a challenge working on those.

I certainly wouldn't take it back to him, but if you're lucky it's just a loose or pinched video cable, and the computer is otherwise okay.

There are, by the way, good independent Mac repair people--some Apple-certified but non-Apple (Apple has a search tool), and others are just good indie shops or individuals that know how to work on Macs. Figuring out the difference between a good place and a clueless idiot can be challenging, of course, which is why you might want to stick to authorized shops, which may be cheaper than Apple direct and have a good chance of knowing what they're doing.

----------

You ninja'd me while I was typing the previous post.

but he ended up installing Mac OS 10.5.8! what an idiot!

That's pertinent info--ignore everything else I said unless it's still not working.

To my knowledge the oldest of the 27" iMacs were in the "late-2009" series, and shipped with 10.6.1. These will NOT, period, work with any OS older than that. I honestly have no idea how the guy even got 10.5 to install--it shouldn't, period--and I'm frankly impressed that it booted at all. Even so, it's not gong to have drivers for the graphics card, and who knows what else--those iMacs were a significant step up from the previous ones in terms of motherboard hardware.

Installing 10.6 with full updates will probably fix most if not all of the post-repair weirdness.

And I'd downgrade the guy from just incompetent to pure moron for the boneheaded OS move.

 

[Chunk1978]

Installing 10.6 with full updates will probably fix most if not all of the post-repair weirdness.

this is great to hear!

 

[Chunk1978]

you were right. installing 10.6.8 fixed all the crazy problems i was having, except for the super drive, which is still selective of which discs it will read. maybe it just needs a cleaning, though i would assume it broke when the HD failed since it only started having read errors after the HD died.

unfortunately, there is one rather annoying problem that seems to be expected when swapping out hard drives on the 27" iMacs: the HD fan spins at full speed all the time. according to the makers of the $30 app "iMac HDD Fan Control":

Since the late 2009 iMacs came out, replacing the hard drive has caused the internal Hard Drive Fan to start running at around 6000RPM. This is due to Apple replacing the external Hard Drive temperature sensor with a proprietary firmware and using the drives internal sensor, replacement drives do not contain the firmware to deliver temperature data on the temperature sensor cable, when the iMac does not receive a good signal from the hard drive it puts the fan at full speed to protect the drive. If you replace your drive your iMac will initially seem fine, but soon the fan will begin to speed up to full speed. Resetting the SMC (System Management Controller) will have a temporary effect, but again the fan will speed up.

the fan is pretty loud! i think my Windows box from 1995 was quieter! any suggestions on that, or am i out another $30?

 

[Makosuke]

the fan is pretty loud! i think my Windows box from 1995 was quieter! any suggestions on that, or am i out another $30?

If you search through the archives of OWC's Blog they did a fair amount of research on the topic, since they do iMac upgrades. I seem to remember there being a way to short two of the extra pins to disable the temp control on that fan to keep it quiet, but you'll have to read through yourself.

 

[Chunk1978]

yes, i've studied the possibilities over the past few days. since i don't have the tools i can't exactly hack the hardware, so a software route would be ideal. on Lion, the only software option that worked was HDD Fan Control, and it worked really great, for an hour (demo period).

i'm currently downgrading to 10.6.8, not a huge fan of those changes in Lion and it still seemed a little buggy at this point (10.7.2) and at this point i wouldn't know what would be causing some errors, my computer or the new OS. hopefully some of the other fan control software will work with Snow Leopard.

 

[Chunk1978]

well, so i thought i was out of the woods, but not so fast.

i installed Lion but downgraded to Snow Leopard, clean install. today i noticed that i'm frequently getting hangs that last under a minute. sometimes i see the spinning beach ball, but other times it just freezes.

reparied permissions, varified disk, ran Onyx. at first trying to open Console.app was a nightmare and would always freeze. after running Onyx it opens lightning fast, as it should. but still, right clicking on the trash in order to get the "empty trash" option seems to always freeze my computer.

any thoughts?

i've attached my console.log

 

[Chunk1978]

damn! i clicked on All Messages in Console.app and every time my computer freezes - still when i right-click on the trash can in the dock, i get this:

11-10-21 9:07:01 PM kernel disk0s2: I/O error.
11-10-21 9:07:01 PM kernel
11-10-21 9:07:01 PM kernel
11-10-21 9:07:01 PM kernel hfs_clonefile: cluster_read failed - 5

that really doesn't look good, especially for a new HD

 

[Makosuke]

I/O errors are exactly what they sound like--something went wrong when reading data off the disk. There are a few things that can cause them:

1) Single corrupt file on disk. Exceedingly unlikely.
2) Directory corruption on the disk related to software, not hardware. This does happen occasionally--I have seen it.
3) Impending/worsening disk hardware failure. Not at all uncommon.
4) Something wrong with the connection to the hard drive. Can be a bad cable--seen this as well, but quite rare.
5) Disk controller hardware failure. Possible, but very rare.

Given the circumstances, it could be #2 due to the previous screw-up (shouldn't be if you reformatted completely) or just really bad luck, but I'd be much more inclined to assume it's an infant hardware failure on the new disk, hence #3. Could also be #4 due to the guy doing shoddy work and pinching a cable or not plugging it in firmly.

Run a Repair Disk in Disk Utility (or fsck via single user mode--which maybe you already did via Onyx) and see if it's able to fix whatever it finds. If it is, keep using the computer, but be conscious to see if it happens again soon.

If problems crop back up again, it's probably a bum disk, though you could have somebody crack it open and check the connection.

Tip: Install SMARTReporter: http://www.corecode.at/smartreporter/

In addition to warning you proactively if the drive is reporting a SMART problem, it will throw up a pop-up in the event of any I/O errors, which will tip you off if it wasn't a one time thing and prevent you wondering what the hang is about.


Aside: My server's relatively new backup drive just flaked out horribly last night, throwing I/O errors all over the place, and I'm left with exactly the same troubleshooting. I suspect hardware failure, but since I've got a franken-mini with externalized internal SATA ports, I'm nervous it's a connection problem with my funky setup, or it could also be the case dying, since it's always been a little weird.

Point being I'm doing the identical troubleshooting by coincidence (unless the Internet is contagious), except in my case it's a lot easier to get to the drive.

 

[Mal]

I've been looking through this whole thread, and I can't see any reason not to have just replaced the hard drive. Just because it was recently replaced, doesn't mean that it can't be failing again, and all the signs are pointing that way. I've see a failing hard drive keep the computer from booting from other sources before, because the system is trying to read the drive no matter what, and if it's struggling, it can refuse to boot from the optical drive too.

jW

 

[Chunk1978]

i booted the computer with Techtool Pro 6 (thumb drive boot, since my computer won't boot with that disc) and did the complete battery of tests. everything passed. it found some favorable changes that could be made with the directory structure (or something like that) so i let it do it's thing. i also installed the Techtool Pro 6 eDrive on a separate partition to make hardware testing easier in the future.

when i restarted everything was faster. right-clicking to empty the trash was also perfect with no freeze. however, after watching a movie i dropped it to the trash and right-clicking to empty the trash once again froze my computer for about a minute. of interest, pressing Command + Shift + Delete empties the trash no problem, it's just that context menu that always freezes the computer.

i just turned on the computer and everything is fine now. no freeze when right-clicking on the trash - for now.

[UPDATE] while the right-click context menu for the trash can didn't freeze the computer, right clicking to close the Console.app did. ugh. wouldn't it be more software related if that simple context menu is the only thing freezing the computer?

 

[Chunk1978]

i'm not sure if this will help, but i'm going to do a 3-pass write-zeros wipe of the hard drive and then reinstall. i read that if there are bad blocks found the drive will get remapped. hopefully the bad blocks are localized and don't eventually spread like a virus.

 

[Makosuke]

i'm not sure if this will help, but i'm going to do a 3-pass write-zeros wipe of the hard drive and then reinstall. i read that if there are bad blocks found the drive will get remapped. hopefully the bad blocks are localized and don't eventually spread like a virus.

While a write-zeros pass is a good way to have the drive remap bad sectors if it was a one-time glitch, nine times out of ten if there are bad sectors now, there will be more soon until eventually the cascade failure eats the whole drive (it will, eventually, fail SMART tests).

With any luck, though, the 3-pass erase will either clean it up or cause it to fail completely--that's a pretty high-stress test for a drive.

And it's the only thing you can really do other than cracking it open and putting a new drive in it. If it works smoothly after the reinstall this time (and do reinstall--just in case it is a software problem, you don't want to be cloning it back), then you're good. If not, probably an infant failure.

And just for reference, there are software things that can hang the Dock, but I/O problems also will. Install SMARTReporter so you'll get warned if that is what's happening.

 

[Chunk1978]

is it possible that the mother board is causing the HDs to fail? or, is it possible that my old HD (which i still have) is perfectly fine and that some other hardware is causing these problems?

it would be unlikely (though certainly not unheard of) that a brand new HD would fail right after it was installed to replace another drive which had died.

how accurate is the S.M.A.R.T. status on Mac OS X?

i was under the impression that if it was the mother board than the entire computer would just not turn on at all. what about a short circuit somewhere? power supply?

 

[chown33]

You forgot to list cables. Bad or intermittent cables can cause HD malfunctions.

And SMART is a function of the HD, not the computer:
http://en.wikipedia.org/wiki/S.M.A.R.T.


If you can't determine the problem definitively, then maybe it's time to take it in for repairs. Guessing and parts-swapping can cause damage. Cables, in particular, are susceptible to damage from multiple insert/remove cycles.

 

[Makosuke]

You forgot to list cables. Bad or intermittent cables can cause HD malfunctions.

That was #4 on my list above.

Interestingly, the internal SATA connector spec only requires withstanding 50 insertion cycles, which gives you an idea of how (potentially) fragile it is. And thus far it's looking like my own I/O errors were due to either a flaky cable (no great shock due to the franken-mini nature of the connection, though it's been pretty reliable otherwise) or the drive managed to correct the problem. I mention the latter because it sounded an awful lot like it was doing some kind of internal routine with a lot of seeks right after I powered it up and before the computer was on; I let it sit for about 10 minutes and it eventually stopped, and hasn't done it since. Could be the case or my imagination (the 5K3000 is very quiet), but could have been it doing remapping.

The latter would make me nervous, but after a reformat and 1.2TB of backup data transfer it hasn't hiccuped so far.