WARNING: The following content might be too graphic for some viewers. If seeing scarred tissue, or the internals of a foreign body might make you ill, you are encouraged to look away now.
Ok, with the wimps gone, we can now go into detail about my recent attempts at recovery data.
First, some history. The drive was stuffed sometime during a stormy night in April 2006. Since that point in time, we’ve been meaning to find the cash to get some photos recovered off the drive.
When I approached several companies for recovery prices, it was rather shocking to hear figures like $3000, $900, $2500, and “upwards from $4000”.
Sure enough, I investigated further what they claim to do, and why they think its worth so much. Sure enough, someone in a jumpsuit with a room that is absolutely clean does the work. Establishing a clean room can’t be cheap, and its not every day a HDD crashes.
I left it sit aside for someday when we have the budget to find someone cheap to do the job of getting some photos and data off the drive.
Recently, as in around 2 weeks ago, I was talking to a friend over MSN about it (came up in conversation), and he mentioned he had recovered a drive. One and one clicked together, and I immediately thought, he probably doesn’t have a clean room in his house, let alone one that those “recovery experts” want to charge us a lot to get a drive dismantled in.
Moving on, the first thing we tried was SpinRite, from GRC. We connected the drive, started it up, and unfortunately, the drive wasn’t detecting.
We tried the freezer trick, to see if that’d make it work. Nope, after around 5 hours of freezing, the drive still enjoyed a good click.
So, next, I decided to investigate some videos found online, at a website, I think it was MyHardDriveDied.com.
Anyway, after reviewing his detailed presentation, I came to the conclusion that, we aren’t in the 85% of software recoveries, so the 10% of recoveries are simple logic board changes. Great, That’s simple, unscrew, chuck a new board on, done.
Luckily, I did my research more before I did that move, and noticed you need a logic board with the same firmware, and a near match as the same drive (also mentioned in the video presentation – another great, legit use of bandwidth).
I remembered, I bought that drive, and a second one with it. I bought two because two machines needed simultaneous upgrades, and it just so happened my method of obtaining computer parts means I try to do a big order, so I discount the shipping as much as possible.
As a result of that, a second drive did exist, and should have came from the same box that this drive came from. My assumption was correct.
So, I immediately got to business in swapping the logic board over after ghosting the 120GB to a 320GB drive (that’s a fair deal for .. stealing a 120GB, outdated drive, right?).
Unfortunately, the one bit that ruined all those few milliseconds where I was ready to scream “I fkn rule” was the first click after pushing the power switch. Dang. So we aren’t in that 10% of logic board recoveries, so we must be in the 4% of head crash recoveries, because the 1% of motor recoveries isn’t me, the motor spins ;).
That 4% seems like a rare number, but anyway, it was certain that drive is gonna come apart sooner or later, so I figure, I’ll go look. The video presentation I viewed mentioned to check the tracks for silver specs, and inspect the dust pillow of the drive. I pulled the drive apart to check it out, and.. well, you see for yourself:
What you can see in the above picture is the outer part of the top platter with a deep scratch in it. The specs you can see on the platter are completely natural, as the drive blows those off as it spins up, that’s the idea of the cushion in the corner, to catch that stuff and hold on to it. The head is the large arm that sits close to the middle chamber that holds the platters.
From there, it was obvious that the head does have to come off, and we still aren’t certain how much damage is done to the underside of the platters (this drive has two platters).
What you can see in this, is the platter, with the head on top of it, and at the bottom sits the second platter.
Anyway, the next step for this drive is to get some sample drives from my MSN friend to play with a head swap to see how that goes, if I get good at that, I’m gonna make a clean box, and pull it apart and do a head swap, and Linux saves the rest of the day again with DD. I tell it to create a full image of the disk to the point of the scratch and we can work with the image at that point which will likely have the data we want recovered.
How’s that for saving several thousand of dollars.
Be aware folks, they aren’t toys, and you shouldn’t attempt this, even though I decided to. You pull a HDD apart, and void warranty at own risk, and any data loss, direct or indirect is on your own hands. Simply, you are highly discouraged from doing any thing mentioned above, and if you do, you do it at your own risk.
The head swap is a little tedious with two platters, you need what I believe is a post it note to keep the heads pushed apart at all times, if they touch, it’s game over. You also need to get the heads back in place in the other drive, to recover the data from, a task in its own when you are working with micro electronics.
Oh, and while I’m posting here, always buy Seagate drives. Nothing’s as tough as a Seagate drive. Even in the video presentation I watched, it was mentioned the highest drives they received issues from were Western Digital, and that completely agrees with my own drive usage.
I’ve had 1 Seagate 80GB drive in my machine for 3 – 4 years or so now, when I built this box back in 2003. That’s how solid they are.
Another Seagate 120GB drive had a bad sector once, was afraid all was lost, started Seatools, waited 30 minutes or so, fixed. Works fine, that happened last year.
Another Seagate 160GB drive just late last year had issues too, it would work, but disappear occasionally, anyway, it powered on for weeks while I waited patiently for an RA number, and eventually sent it back, and replacement works fine. I didn’t lose any data from any of the Seagate drives described above.
This Western Digital is the first Western Digital drive, and.. well, it has one storm go buy and bombs out, in a room that had around 8 or more HDD’s running at the same time.
My server here as a 40GB WD, and a 80GB Seagate, looking at SpeedFan’s SMART viewer, the WD has low rating..
The Seagate drives all have very high ratings, and have very long uptimes. The only time my entire network is down is when a power outage occurs, or when I am doing something that means taking machines down.
I’ve become a lot greener with electricity too, I’ve moved a few machines into virtual machines on the one server and now each machine has a real purpose requiring being turned on, as a result our power usage seems to be realigning itself to a cheaper scale, and the $400 power bills should be gone (it’s nearly that time again too).
Anyway, that’s what a HDD is like for you.