Recovering a dead hard drive

[shmokes]

I've been having lousy luck lately with electronics. Anybody know of some software that can potentially recover a hard drive that won't mount at all or show up in windows Disk Management? The hard drive does get power and spin, but it sounds like something might be wonky mechanically (read: I think this is probably a lost cause) because from the sound of it, it's trying over and over again to do something--like, it starts spinning up, then stops and starts again, then stops and starts again. Or maybe the arm is stuck. I don't know. But the hard drive is at least doing something, which gives it a better chance than if it wasn't. 

It's not a data issue. It's part of a RAID 5, so I can wipe the drive clean and I still won't lose any data. Also, the data on it is all backed up elsewhere. I'd just like to get it going and not have to replace the thing, especially right now cos hard drives are crazy expensive since all the floods hit Thailand.

[CheffoJeffo]

It's not a data issue. It's part of a RAID 5, so I can wipe the drive clean and I still won't lose any data.

Until you try and use it ... 

[shmokes]

Right . . . well, even if the array failed to rebuild, like I say, that's not an issue for me. But I'd like to save the drive if I can.

[CheffoJeffo]

Right . . . well, even if the array failed to rebuild, like I say, that's not an issue for me. But I'd like to save the drive if I can.

Right, but a failed drive is a failed drive ... hence, it has failed and shouldn't be trusted with data, which presumably is worth far more than the drive.

[Dervacumen]

Is it by chance a Seagate 7200.11?  If so, I might possibly help.  Something about some bad firmware setting the busy flag and not releasing it.  I built a cable, sent some commands to it via hyperterminal, and saved a 500GB drive that way.

[shmokes]

Nope. Hitachi Deskstar 500 GB (I know ).

[shmokes]

Thanks for the suggestions. I'll see if any of those work.

[Xiaou2]

Cheffo is Right.

 What you are hearing is the "Click Of Death".

 Theres no way to save the drive.  The read heads are Toast.

 Imagine this... your driving down the road at 100mph, and you run smack into a 4ft diameter boulder.   Your car is trashed.   No software will fix this.

 The HD heads are smaller than the a human hair.  If the drive head hits a tiny spec of dust at high rpm.. its finished.

 Other failures such as strong vibration (stomping on the floor, crazy subwoofer..etc), can cause the heads to Crash into the high speed spinning platters.  Again, trashing the heads.

 Software recovery only works on things like drives with bad sectors, corrupted data, etc.
 

[SavannahLion]

Oh wait? How the Hell did I miss that?! I hate being sick. You're trying to recover the drive itself? No no no the tools I mention are to recover data from a failing drive, not to recover the drive itself. If you're using a RAID and the drive dies, turn the HDD into a clock because that's all it's good for.

I'm deleting my other post. It's way off base....

[CheffoJeffo]

Cheffo is Right.

And that is how you know that it isn't just a matter of opinion.

Xiaou2 is right.

[shmokes]

Careful . . . I see a veiled insult in that last post 

It's actually not the click of death, per se. Perhaps it may as well be, but what I'm hearing is the hard drive spin up for a good five seconds or more then spin back down, then spin up again, etc.  As for the value of the data, that's simply not a concern. Like I said, this is a RAID 5 array (one layer of data protection) that is regularly backed up. If I put this hard drive back into production and it fails, I'm in the same situation I'm in now. I just have to replace it and rebuild the array, no harm done. If I replace it and the array fails to rebuild, I still haven't lost anything because I have all the data backed up, so I can just recreate the array, restore the backup, and the problem is solved. Obviously it's possible that I put the drive back into production and then it fails, then the array won't rebuild, then the backup will have gone corrupt. A perfect storm. But you can't live your life taking every possible measure to protect yourself from even the slightest risk. Or at the very least, the data on this drive is not valuable enough to warrant such measures.

Anyway, it's probably neither here nor there. Like I mentioned in the OP, it's probably a mechanical failure. But if something as simple as a corrupt FAT could cause similar behavior I wanted to pursue it.

[Xiaou2]

Run a hard-drive test utility by the MFG of the drive.  Theres usually a short test, and a Really long test.

[SNAAKE]

time to invest on a SSD drive

[CheffoJeffo]

Careful . . . I see a veiled insult in that last post 

 

[lilshawn]

Nope. Hitachi DEATHstar 500 GB (I know ).

^^fixt^^

[Thenasty]

I did an experiment by buying the SAME DRIVE/MODEL. Open it up and swap the PLATERS and........................................................














































it did not WORK  

Now I got 2 drives toasted  

well if it work, it would have been worth it but I knew the risk.

[Samstag]

I did an experiment by buying the SAME DRIVE/MODEL. Open it up and swap the PLATERS and........................................................

it did not WORK  

Now I got 2 drives toasted  

well if it work, it would have been worth it but I knew the risk.

Did you try swapping the controller board first?  It's a hell of a lot easier and won't ruin the replacement.

[Thenasty]

yea, that was the first part that came off.

[Samstag]

When that doesn't work, it's usually time to give up.

[lilshawn]

when you separate the motor from the platter(s) you misalign the sectors. opening a drive and allowing any dust into it

Imagine this... your driving down the road at 100mph, and you run smack into a 4ft diameter boulder.

yes this... with stuff this small, a few skin cells in a piece of dust might as well be a boulder.

when they produce the drive initially they do a special formatting to the drive to prepare it for low level formatting. that way the heads can track the sectors properly. It's essentially like a timing signal.

even if you marked the disk platters and attempted to align them again, the sectors on modern hard disks are so small a few nanometers (less than maybe 10) of drift would be all that would be needed to render data difficult to read.

if you wished to retrieve the info (and in this case you wouldn't since it's just a portion of raid) and you have a bad motor, you need to have it professionally done so they can re align the sectors again. they use a special magview machine to look at the disk magnetic structure and re align the platter again. when used in data recovery, they use a special seperate motor and head structures for each platter that they can synchronise several platters separately and read the data synchronously. once the data has been read, the original is discarded.

i used to have a photo of a platter that had a head crash due to a rough bang. the soft aluminium essentially had 4 tiny little spots (the four corners of the ceramic head structure) then, about 1/2 inch away was a scrape mark the width of the head that continued about 3/4 of the way around the platter. every time the head hit the spot of the initial crash (4 tiny spots) the head would launch off the surface of the disk and crash land of the platter a half inch away. then scrrrrrrraaaape until the head stabilized and the air gap was restored again. until it hit the spot again..... 7200 times per minute. i managed to get the unit shut down in a few seconds but the damage was done. if it was allowed to continue the drive would certainly wear down the head and platter to nothing.