Oddest crash/no startup I've come across, please help

[jimbles]

The basics first: 24" 2008 iMac, 4GB RAM, OS 10.6.8, all updates, running second monitor via mini-DVI

So, I've experienced an odd crash that I've seen similar before, but not this bad off. I was copying a family vacation DVD to mp4 using Handbrake, and left the room to watch a movie. My iMac is set up to go to sleep after about 1.5 hours, and it did so by the time I returned to the room. I pressed a key to turn it back on, had the password prompt over the black screen pop up, I typed in my password, and... waited. I could hear the DVD drive spinning, I could hear the hard drive making a spirited effort, and I could hear the fans. The beachball popped up, and no typed letters were being written in the password prompt. After about 10 minutes of this, the screen went off altogether and could not be brought to a wake. It seems as though it just powered down entirely. I decided to restart.

During restart I ejected the DVD. The machine whirred, and eventually the dreaded flashing missing folder icon came up. This had happened before, ages ago, under similar circumstances (forced reboot), so I powered off, disconnected the power cable, let it sit a while, and turned it back on (which worked prior). This time, nothing, still flashing folder.

Reset again, tried to do PRAM, but it didn't even respond to PRAM by doing the usual restart operation, it just went to the flashing folder.

Inserted OSX install disc, held down option, and the main drive partition didn't even show up. Went into OSX install disc's Disk Utility, checked, and again saw no OSX partition.

I am convinced that the hard drive itself isn't fried. I'm also mildly certain that the OS itself is still intact. My reasoning is, if the file structure itself vanished, or the drive was fried beyond working, how could it display the login prompt? Somehow, though, something must have triggered something massively bad where it's in perpetual hang and can't get unstuck (boot order? something with improper permissions? failed write of an important file?). Any suggestions? I have a many months old Time Machine save, and have all of my important work documents stored separately elsewhere so a complete reinstall isn't out of the question, but is there anything anyone else can suggest that might fix this?

Thanks, guys.

 

[velocityg4]

No big mystery sounds like the hard drive is dead. The login prompt showed when waking from sleep since it resided within RAM it wasn't able to go beyond that since all OS files do not reside within RAM. No process you run is going to suddenly damage a hard drive (besides a botched firmware update). Most of the time when they fail it is sudden and complete.

The only thing that is going to fix it is replacing the hard drive and doing a fresh install. Even if you could possibly get the old drive working to do a fresh install I would not trust it for a second. Once a hard drive has any problems I replace it. They are cheap.

Stop trying to boot off that hard drive. Every second the drive is spinning decreases the chances of recovering data off of it. If you don't care about the data I'd still stop fiddling. The drive is no longer trustworthy so you are just wasting time on it.

 

[1934hotrod]

I agree, boot the machine from the CD open the utility ... see the drive, if answer in NO, drive has failed. Real simple happens all the time.

 

[jimbles]

Interesting that you guys would say that it's a lost cause and needs to be minimum wiped, likely replaced... I decided to unplug the iMac a bit longer this time (about 15 minutes), then plugged it back in, pressed the power button, prayed, and... Success! It took a good while for it to come to the grey screen with the Apple logo on it, but eventually it did.

And the weirdest thing? It went RIGHT to the woken from sleep login prompt. Not the fresh boot login, I'm saying the sleep login. Then, right straight to the desktop, with Handbrake running and a message saying my encode was finished. This is after being powered off, unplugged twice, booted from the install DVD, booted raw numerous times... I can't explain it any other way than a complete and utter anomaly or super unique OSX bug. I have a hunch that my first instinct was right, that somehow something was overwritten in terms of which drive to boot from. Or it could have been a myriad of other issues, but after running checks it looks like the drive is completely fine, sectors are free of faults, etc. So let this be a lesson: don't jump first to wiping everything with hellfire if it doesn't work the first time.

 

[monokakata]

How about looking into the system logs via Console? There might be useful clues there.

Unexplained failure and unexplained recovery should be looked into as carefully as possible.

It would be good to find something pointing towards a cause.

 

[jimbles]

Yeah, I've been scouring the logs for quite a while now, looking for hangs and messages that fit in the timeline. From system.log:

Feb 9 03:49:01 iMac loginwindow[31]: loginwindow SleepWakeCallback WILL sleep
Feb 9 03:49:02 iMac configd[14]: PMConnection mDNSResponder com.apple.powermanagement.applicationresponse.slowresponse 232 ms
Feb 9 03:49:02 iMac configd[14]: PMConnection IPConfiguration com.apple.powermanagement.applicationresponse.slowresponse 257 ms
Feb 9 05:05:10 localhost com.apple.launchd[1]: *** launchd[1] has started up. ***
Feb 9 05:05:18 localhost DirectoryService[15]: Improper shutdown detected
Feb 9 05:05:25 localhost fseventsd[35]: event logs in /.fseventsd out of sync with volume. destroying old logs. (44357 17 48940)
Feb 9 05:05:27 localhost bootlog[42]: BOOT_TIME: 1360411508 0

The 5:05 mark signifies where it booted properly. Noticed how I used the term 'booted', because despite returning me to the exact spot where I had last had the computer set before it went to sleep (with Handbrake running, a message for it displayed, etc) and bringing me immediately to the post-sleep login screen, it registers as having gone through a full boot in the console.

The 03:49 shows a pretty regular sleep process, which would be the last thing before everything went belly up temporarily.

So evidently there's absolutely nothing in terms of a record of me restarting the machine those several other times and getting the flashing folder. I don't even see a record of it ejecting the DVD that I was working on, because that happened during the no-boot fiasco. It's just normal operations until time of sleep, then the standard boot message, then the standard messages for post-sleep login prompt later on down.

It's absolutely one of the weirdest anomalies I've ever witnessed in all my years running OSX, as though no problem ever existed. The volume just plumb didn't mount originally, it would seem, and thus there's no record of any major errors. And when it did detect the partition and boot, it didn't revert to default boot settings, but rather jumped to post-sleep settings.

Has anybody ever seen something like this?

 

[chown33]

So evidently there's absolutely nothing in terms of a record of me restarting the machine those several other times and getting the flashing folder.

If the disk wasn't working at the time, as signified by the flashing folder, then how could anything be written to the log files? Log files reside on disk. The disk must be writable for logs to be written.

I don't even see a record of it ejecting the DVD that I was working on, because that happened during the no-boot fiasco. It's just normal operations until time of sleep, then the standard boot message, then the standard messages for post-sleep login prompt later on down.

Same thing applies here: no writable disk, ergo no writable log files.

It's absolutely one of the weirdest anomalies I've ever witnessed in all my years running OSX, as though no problem ever existed. The volume just plumb didn't mount originally, it would seem, and thus there's no record of any major errors. And when it did detect the partition and boot, it didn't revert to default boot settings, but rather jumped to post-sleep settings.

Has anybody ever seen something like this?

I haven't seen anything quite like that, but it's not beyond possibility.

First, if the computer goes to "deep sleep" (also called "hibernate") then it writes its current state to disk. At the next wakeup, that state is read in from disk, resuming almost literally right where it left off. The whole reason for deep sleep is it preserves the current memory state in a recoverable way on the disk, so even if power is completely lost (typically a laptop battery dying), the computer won't lose anything.

Second, I strongly suspect your hard disk is failing. Impending disk failures have these types of symptoms: extremely long times to read or write, intermittent and unpredictable delays (due to long read or write times), and apparent recovery after being turned off and allowed to cool to ambient temperature.

If I were you, I'd shut the machine down, buy a replacement HD, and make a complete backup. Leaving the machine on can lead to complete disk failure. Honestly, so can turning it off, but in the several times I've had HDs fail with similar symptoms, turnning off was better than leaving them on.

By the way, none of the other replies said to wipe the hard drive. They said to replace it.


From your OP:

My reasoning is, if the file structure itself vanished, or the drive was fried beyond working, how could it display the login prompt?

The code for doing the login prompt was probably already in RAM, and didn't need to be read from disk.

Imagine the computer is going into deep sleep, so it writes all of RAM to disk. The code for managing deep sleep is probably in a single library containing everything needed for deep sleep and subsequent wakeup. Thus, to go into deep sleep, the wakeup code is loaded, even though it isn't used yet.

After the RAM is written to disk, the disk is spun down and turned off. You wake the computer. The login dialog is presented (previously in RAM) and you enter the password. The disk is turned on again, but now it's not responding (extremely long read times). The computer seems to hang. You turn it off, losing RAM but not affecting the disk, which remains unresponsive (reads or writes).

As long as the disk remains unresponsive, nothing can be read or written from it. There's still a deep-sleep file on the disk, but the wakeup attempt failed, so it's still there and ready for use. So when the disk starts working again, the computer reloads that state and continues where it left off. Remember, the saved date for deep sleep is on the unresponsive hard disk. Since that disk was unresponsive, there's no way to delete or override that data, not even intermediate boots from DVD.

At best, that's a "simplified story", but as I said, it's not beyond possibility.

 

[jimbles]

Yeah, in terms of price it's not exactly financially punishing to order a new drive for this iMac. I've been toying with the idea for a while, as a bump up from the measly stock 250GB to a terabyte would be fantastic. I'll put an order in this week and have a nice weekend project coming up.

The scenario that you typed up certainly seems plausible. Still, I hate the fact that there wasn't a comprehensive error log write up on the system and that everything is coming up negative in terms of checking the disk for sector errors, and I've seen no other signs of impending hardware failure aside from this incident (data read/write speeds seem fine for a 5400rpm drive, all beachballs I've had stem from RAM usage, no patterned clicking noises, never had to shut down due to heat, etc). It's like seeing smoke, and knowing a place where a fire could happen, but not being able to know where else the fire may have spread to.

 

[Fishrrman]

A few thoughts:

Next time you do something like this (anything which requires a long, unattended task which might interfere with "sleep"), set sleep to "never" first.

Also, with a 2008-vintage Mac, I'll reckon that the hard drive is SEVERLY fragmented. I would boot "externally" (i.e., from a cloned backup drive), and use a defragmenter/optimatization utility on it (such as iDefrag or Drive Genius).

Disregard what others may say about "the Mac doesn't need defragmentation". Try it, and observe the results...

 

[chown33]

The scenario that you typed up certainly seems plausible. Still, I hate the fact that there wasn't a comprehensive error log write up on the system ...

The hard disk was unavailable to write any logs to. If there's nothing to write a log to, I'm not sure what you're expecting to happen. Logging isn't magic; it needs a medium.

... and that everything is coming up negative in terms of checking the disk for sector errors, and I've seen no other signs of impending hardware failure aside from this incident (data read/write speeds seem fine for a 5400rpm drive, all beachballs I've had stem from RAM usage, no patterned clicking noises, never had to shut down due to heat, etc).

Disks fail in different ways.

I've personally experienced 3 HD failures on my Macs over the years. Of those, one was a complete and instant failure, with no warning signs at all. It occurred while the machine was running, and the disk was powered on. It just stopped working and never came back.

The other 2 failures had warning signs of extremely long read times. In one case, these gradually became more common, to the point where I finally recognized the symptom. Before that recognition, it was sporadic enough that I had no explanation, and since it went away again for a week or two, I wasn't seeing a consistent pattern. In the other case, the first appearance of long-reads were quite common, and I immediately recognized the symptom. In none of the failures did I see SMART status indicating an impending failure. And never a clicking noise.

It's like seeing smoke, and knowing a place where a fire could happen, but not being able to know where else the fire may have spread to.

Sometimes the safe thing is to move to a new house. The $100 or so cost for a replacement drive is insignificant compared to the time and effort of recovery from a failed drive, in my opinion. Your cost/risk evaluation may differ from mine.

 

[snberk103]

I'm with chown33 here.

I've been working with computers since the mid-1980s. In my experience the general rule of thumb is that sporadic errors are usually hardware issues, while software issues produce consistent errors. This sounds like a sporadic hardware issue - and points at a hard-drive problem.

A couple of weeks back our nephew was visiting with his Mac Book Pro - newish. It was throwing up weird communication errors, and he was convinced it was software and/or file system issues and wanted to install all of these file optimizing applications. I figured it was the HDD.

So we made a backup, and reinstalled the OS. And we starting getting long write times, long boot ups, as you're describing. Long story short, I had AppleCare on the line (I like our nephew that much - and he was pretty stressed) - and they were having me go through their usual hoops... when the HDD stopped responding. No sounds... no other warnings... it just stopped responding, so our nephew got a new HDD courtesy of Apple. And I got a reputation as a miracle worker.

Do a cloned backup. Reinstall the OS ... that may stress the HDD enough that it finally consistently fails. Get a new HDD and install it. Or at least have it handy to install.

Actually - save a step - buy the new HDD and clone the old HDD to the new one. You now have a backup. Swap drives and the new drive should just work, and the old drive now becomes the backup in case the new HDD has issues.

 

[jimbles]

I appreciate all the responses. I can see you guys are having fun addressing this concern, you all have quite the knowledgeable pool to draw from!

Anyways, new drive was order from Newegg, should be here this week. Not really looking forward to ripping apart the iMac, but I've torn apart laptops before and am not too concerned.

To the point: Knowing that it's all on 'borrowed time,' I booted the (working) drive again and grabbed some small files that I needed and hadn't backed up yet. It worked, but after a while the drive slowed to a crawl, and perma-froze. Upon rebooting... yep! Same missing folder problem. It's toast, you guys were absolutely right.

Here's the thing, though: I don't want to stick a new hard drive into a machine that may have larger problems, possibly one that could be frying the drive to begin with. One thing that I noted while all these problems were happening was the high level of heat building up quickly, far more than even during power use sessions. Naturally I'm worried that I'll just be sending another drive to the slaughter.

So I tried to run Apple Hardware Test. I have my original Leopard install DVDs, and popped in the one with the AHT label, held down D at start up, and... Nothing. It went to the blinking drive folder. Since my current install on my drive is 10.6, I figured (against logic, as the drive isn't even registering to begin with), maybe I'll need to run the AHT that came with the retail 10.6 DVD that I got. Nope, no go. Nothing.

Am I missing something? I was under the impression that AHT was supposed to run off the DVD independent of a functioning hard drive? Thoughts? Sorry to shift gears in questioning, by the way.

 

[jimbles]

Forgive the bump, but I've started reading up on others who have had a similar flashing folder issue on their 2008 iMacs. Though some seem to get by on fixing their hard drives, it seems like many who were still under the Apple Care warranty and submitted their computers for fixing received news of failed logic boards that needed replacing... Quite the scary (and highly pricey) situation on that one.

Could my issue fall under a logic board failure? Is it symptomatic? My optical drive appears to run just fine, which seems to somewhat fly in the face of logic board failure... Really wish I could figure out why AHT isn't working and maybe see if there's a larger issue at hand.

Guess I shall find out tomorrow when my drive arrives (I hope!) and I swap it out. I'll update and let everyone know how it turns out.

 

[Brian Y]

Forgive the bump, but I've started reading up on others who have had a similar flashing folder issue on their 2008 iMacs. Though some seem to get by on fixing their hard drives, it seems like many who were still under the Apple Care warranty and submitted their computers for fixing received news of failed logic boards that needed replacing... Quite the scary (and highly pricey) situation on that one.

Could my issue fall under a logic board failure? Is it symptomatic? My optical drive appears to run just fine, which seems to somewhat fly in the face of logic board failure... Really wish I could figure out why AHT isn't working and maybe see if there's a larger issue at hand.

Guess I shall find out tomorrow when my drive arrives (I hope!) and I swap it out. I'll update and let everyone know how it turns out.

You can never rule out other problems, but I'm 99.999% certain it's just the hard drive. What you've described is a typical hard drive failure.

It's not that bad replacing the HDD - make sure you have a can of compressed air with you to a) give the fans a good clean, it'll be dusty after 4 years! and b) get any dust out from between the glass and the LCD panel when you put it back.

Couple of things worth bearing in mind with the 2008 ones:

- When removing the glass, use two suction cups at *diagonally opposite* corners, and pull them together, and straight off. Don't pull at an angle.

- Remember to discharge the PSU - it's not enclosed, and the caps can still give you a nasty shock even after pulling the plug

- Be careful of the small cables, such as the microphone cable (which runs to the right of the webcam) - it will need to be disconnected *while* removing the front bezel, but the connectors are taped together (so you need 4 hands, or good balance ).

 

[chown33]

Could my issue fall under a logic board failure? Is it symptomatic? My optical drive appears to run just fine, which seems to somewhat fly in the face of logic board failure... Really wish I could figure out why AHT isn't working and maybe see if there's a larger issue at hand.

The problem could be a logic board failure, but I see no specific evidence it is. I've seen more disk failures, and no symptom yet posted suggests it's NOT a disk failure. Nor does any posted symptom scream "logic board failure".

In that computer (2008 iMac) I'm pretty sure the HD and the optical drive are on different interfaces. If they are, then one interface can fail (e.g. the connector where the cable goes to the HD) while the other interface and connector work fine. So while it's possible for the logic board to be at fault, the simplest (and probably cheapest) way to find out is to replace the HD. I assume you read the return policy for the HD vendor, in case the replacement HD exhibits the same symptoms.

AHT can't diagnose a problem if the disk isn't failing at the time the diagnostic is running. Since the HD failure is intermittent, the disk could appear to be working when AHT is running. I'm not even sure AHT would be able to determine that a long read-time should be considered a failure.

 

[jimbles]

You can never rule out other problems, but I'm 99.999% certain it's just the hard drive. What you've described is a typical hard drive failure.

It's not that bad replacing the HDD - make sure you have a can of compressed air with you to a) give the fans a good clean, it'll be dusty after 4 years! and b) get any dust out from between the glass and the LCD panel when you put it back.

Couple of things worth bearing in mind with the 2008 ones:

- When removing the glass, use two suction cups at *diagonally opposite* corners, and pull them together, and straight off. Don't pull at an angle.

- Remember to discharge the PSU - it's not enclosed, and the caps can still give you a nasty shock even after pulling the plug

- Be careful of the small cables, such as the microphone cable (which runs to the right of the webcam) - it will need to be disconnected *while* removing the front bezel, but the connectors are taped together (so you need 4 hands, or good balance ).

Thanks for the tips - I've been browsing the various iFixit iMac repair guides and breakdowns, trying to psyche myself up to the task. Haven't seen the PSU tip yet, so that's good to know. Typically when I had to move my iMac around from a different state I would unplug and hold down the power button for a few seconds to discharge the built up electricity, something I learned after being shocked while trying to fix a DVD player, haha.

I have a whole slew of supplies that I bought recently: Compressed air, clean white cotton gloves, plastic spudger, 2 medium sized suction cups, one of those soft cloth wipes for eyeglasses... I should also have access to another set of helping hands, so it should be all good.

The problem could be a logic board failure, but I see no specific evidence it is. I've seen more disk failures, and no symptom yet posted suggests it's NOT a disk failure. Nor does any posted symptom scream "logic board failure".

In that computer (2008 iMac) I'm pretty sure the HD and the optical drive are on different interfaces. If they are, then one interface can fail (e.g. the connector where the cable goes to the HD) while the other interface and connector work fine. So while it's possible for the logic board to be at fault, the simplest (and probably cheapest) way to find out is to replace the HD. I assume you read the return policy for the HD vendor, in case the replacement HD exhibits the same symptoms.

AHT can't diagnose a problem if the disk isn't failing at the time the diagnostic is running. Since the HD failure is intermittent, the disk could appear to be working when AHT is running. I'm not even sure AHT would be able to determine that a long read-time should be considered a failure.

Yeah, there's nothing really wrong with having another spare drive sitting around, so if it is a larger (and more costly) issue at hand I won't be too frazzled about ordering the HD. I am hoping that this clears it up, though - I imagine that if it doesn't, the next logical step would be to ensure there's nothing wrong with the SATA connector itself, meaning another purchase and another few days of downtime while it ships.

About AHT - The thing I'm having a hard time with is the fact that AHT isn't even loading at all, period. Wouldn't it still load from the install DVD, independent of a failing HD, if I use the D key before the grey screen loads? Part of me is thinking that the boot loading sequence and the hard drive hang is interfering with it loading, but at the same time I can still access the other functions of the install DVDs, such as Disk Utility, so that sorta contradicts that to me. This is just such a myriad of mysteries with this issue.

 

[jimbles]

Okay, update time: I'm currently typing this post... on my iMac! It was a rousing success. monokakata was kind enough to PM me a link to a tutorial video that I used in conjunction with both the iFixit guides as well as the video tutorial featured on OWC's website.

The replacement of the old WD 320GB stock drive with a fresh, SATA III 7200 rpm WD Black 500GB drive went off pretty smoothly, save for a few bumps here and there. If anyone with a similar issue finds this thread through the forum search, here's some things that I experienced while tearing the iMac apart:

-Up until the part where you need to start disconnecting wires, it's a cake walk. Glass comes off with barely a hitch (I've sen recommendations for putting 2 suction cups diagonally on corners and a different tutorial saying to put 2 cups at the top only... I went with the top, as it allowed me to pull down smoothly and rest the bottom of the glass on the edge of the aluminum depression), the bezel is easy with just a handful of Torx screws, etc. Have everything prepared ahead of time and you'll work through it smoothly.

-I would recommend getting a soft bed sheet to rest the glass on once you've removed it, and if you have pets in the house like me, run a lint roller over the surface of the sheets a few times immediately prior to touch down. Saves some hassle later.

-I chose to wear white cotton gloves, the semi-disposable kind that people working at a print shop would wear. Again, if pets are in the house, lint roll your gloved hands prior to handling the glass as well as the LCD screen.

-Work on a LARGE open surface. I worked on a dining room table made to seat 8, and I still felt choked for room with all the supplies laid out. Remember that you'll not only have your tools out, but you'll need to lay out the full aluminum bezel, the LCD, and the computer itself, with ample room to spin it about as needed to fish out tricky wires. Space is key.

-Some of those wires were TRICKY to get out. In the tutorials you see them pluck right out of their connectors, but in my case they were extremely tough to get out. My guess is either a World Strongman competitor was taking them apart in the tutorials or they are using machines that have been loosened up from numerous breakdowns in preparation for filming. Be patient, don't lose your cool, and don't yank too hard on the wires themselves. Grow your nails out if you can... I'm a habitual nail-biter and it was crappy trying to get a grip on those tiny connector heads with bare skin.

-There was a surprising amount of tape on the connectors, which the tutorials did not show (likely because they had already peeled it off during a test breakdown prior to filming). It's fine to peel it off, though I chose to leave the tape on the wires connecting the hard drive because it helped to bunch them up neatly. The tape around the microphone connector head was discarded.

-The aluminum bezel came off extremely easily, but putting it back on was a tough one at first. A top priority is making sure that the iSight cable up top isn't crimped at all, so watch out for that. The other thing worth noting is that you need to start putting it back on from THE BOTTOM, because there is a barely noticeable plastic lip that runs the entire length of the bottom of the back black section, helping to secure the bezel. Get the bezel under the lip at the bottom, then work up, and you'll be fine.

-I live in Arizona, so you can only imagine the level of dust I found inside the computer when I opened it. Have a can of compressed air ready, and after you've removed the glass and LCD, take the computer body far, FAR away and give it a good blast of air. Target the bottom fans, the bottom perforated intake holes, and the perforated holes towards the top where the air blows out. Then give it a general dusting everywhere else.

-The 2008 iMac uses a stock SATA II drive, which has a 3.0Gb/s threshold. SATA II are becoming increasingly obsolete and, frankly, harder to find a decent selection of with quality modern features, so I went with a SATA III WD Black drive. I also look at it as an investment for later, should I decide to take the drive out after decommissioning the iMac, such that I can have access to that higher speed output in another machine capable of utilizing it. To ensure no problems with the drive, I put a mini sized hard drive jumper on the new drive to throttle its speeds down to 3.0Gb/s. It may be overkill, but I couldn't get a definitive answer as to whether my particular iMac would go bananas with a SATA III in it, so I figured I'd save the hassle of having to open the iMac up again and just gimped the speeds while I could (it's not like it could have SATA III speeds anyways). System Profiler is reading it as SATA II (3.0Gb/s), so it looks like it's mission accomplished. Your drive manufacturer should have a guide as to what pins to put the jumpers on to achieve this. With my particular SATA III WD Black (WD5003AZEX), the jumper went on pins 5-6.

-I only noticed this after I had pieced everything back together: The stock WD drive inside the iMac had the fuzzy foam stuck on it... on both sides! One piece of foam covers the heat sensor (and was thoroughly fused to the drive itself; start prying it off right at the sensor, where there's some give, not at an odd corner, because you'll just rip it off and leave a stuck on strip that you'll have to peel off), the other piece of foam is on the back side of the drive, a little tiny square about 3/4" big. I missed the back square entirely, and now my current drive is without it. Hopefully no issues arise from it, but I can't imagine something as inconsequential as that will do much of anything. It seemed more just like padding than anything of true higher function.

-The connector bit tucked underneath the mylar tape on the backside of the LCD was a BEAST to disconnect. I had my brother assisting me, him holding the LCD upright at an angle, me underneath getting at the connector. My advice is, have an assistant, and have them match your force of pull but going in the opposite direction, giving more stable leverage. He pushes the LCD up, you pull the cable down. Do the opposite when reapplying the connector (assistant pushes down, you push up). The only way I could see this working if you are doing this solo is to completely forego the disconnecting of the connector under the mylar (which monokakata suggested to me in his PM), or having one hand on top of the LCD screen and one hand pushing in the connector, which would be insane and could ruin your LCD with all the pressure. Seriously, ask for help. Unless you don't mind propping up the LCD the entire time with your forearm as you work to undo the hard drive, which could work.

I think that's pretty much it for tips regarding areas where I had troubles/made observations I hadn't seen before. I'll have this thread on notify mode with my email, so if 5 years from now someone finds it in an hour of desperation and has a question, I'd be happy to answer.

Oh yeah, one last thing: After I got the old drive out and the new one in, I ran the old drive through my Rosewill USB dock. Guess what? It just did the start up whine, a few patterned clicks, a few repeats, then stopped. No mounting at all. You guys were spot on in saying it was a dying drive, and this pretty much proves it. I currently have it in the freezer, wrapped in clingwrap and tucked in 2 ziplock bags. We'll see if I can't get it to do... something tomorrow. There's 1 or 2 files I'd like to get access to, after all.

If anyone has frozen drive tips, by all means, post them.

 

[blueroom]

Do you have a backup? Time Machine would be ideal. Ahh no current backup I see in your last post.

Time to bury that old drive, 5 years is typical for mechanical drives and sooner or later most folks will have a drive crash. Recovering anything off that drive seems unlikely.

PS get a backup and at the least store your important stuff somewhere i.e. DropBox, Box, Skydrive...

 

[jimbles]

Do you have a backup? Time Machine would be ideal. Ahh no current backup I see in your last post.

Time to bury that old drive, 5 years is typical for mechanical drives and sooner or later most folks will have a drive crash. Recovering anything off that drive seems unlikely.

PS get a backup and at the least store your important stuff somewhere i.e. DropBox, Box, Skydrive...

I actually did have a Time Machine back up, done both a week prior to the first crash and right in the middle of a lull in between crashes when the drive miraculously worked. I only lost a few things that I had saved the day of the final crash, so it wasn't really too big of a thing.

I do use dropbox and google drive with my company work files, but it's typically just final products that are uploaded and rarely process work, so that's where I sort of goofed. Basically I just lost a day of progress on something I was working on during the final freeze, which I'm not really too concerned over; after all, in graphic design, once you make something and have to make it again, you'll do it 50 times faster, hahaha.

Freezing the drive though is more of an experiment, something I've always wanted to do but never had a borked drive to do so with. My only concern is condensation from the thawed drive running into my USB drive dock, but it's a small concern anyways, as I would have it going for barely 2 minutes if it fails to load anyways.

 

[blueroom]

I remember the freezer trick from ancient IDE drives, never seen it work though.

SpinRite can hammer away a drive for days trying to recover / remap sectors.

 

[snberk103]

Good News! And thanks for the update.... That disassembly description makes me glad I have a Mac Pro...Yikes!

 

[jimbles]

I remember the freezer trick from ancient IDE drives, never seen it work though.

SpinRite can hammer away a drive for days trying to recover / remap sectors.

The only issue is that it won't mount as is, period. Even in a USB dock the drive is non-mounting, so traditional data extraction is going out the window, haha. Guess I'll have to report back in a few hours once it's sufficiently chilled (I've seen a recommendation of 12 hours, I'll let it chill a bit longer. Maybe after my work shift, or something) and i have a chance to see it pass or fail. I will say this though: In full disclosure, I may have run it too much after its last bout of death throes. I just couldn't leave well enough alone, and may have driven it as far into the ground as possible, hoping against all hope that it would repeat a miracle boot pre-disassembly. Ah well.

Good News! And thanks for the update.... That disassembly description makes me glad I have a Mac Pro...Yikes!

Haha, for sure. Lesson learned: If you're a tinkerer or plan on having something well past the end of a warranty, avoid all-in-ones (man, I can't even imagine what sort of pain this would have been if I were working on the new iMacs, what with their adhesive-stuck screens and all...). With how much progress the Minis have made these last few iterations, what with i7 quad cores and 2x drive slot capabilities, they look pretty nice for my moderate intensity graphic design needs. Heck, the 16GB RAM alone dwarfs my iMacs piddly 6GB limit. If the next version in the fall has something better than just anemic on-board shared GPU, like say expanded GPU via Thunderbolt (unlikely), I'm 100% down. Shame I'd have to fall back upon my ugly Asus monitors and forego the high priced IPS Apple stuff, but that's life on a budget.

 

[jimbles]

Update on the frozen drive experiment: Pretty much a complete no-go. Froze it for 24+ hours, well protected from moisture, and there wasn't much of any change. The only distinct difference this time around was that the drive sounded quite alright, none of the patterned clicking noises, but after its initial powering-up and a few whirls, there was a loud and distinct clicking noise, almost like a 'thunk' that basically killed its sequences.

Well, now I have a tear-down specimen to examine and have fun with by learning the internal mechanisms of a drive. Who knows, I may even spot an issue that I can fix along the way. Silver lining, you know.