Hard Drive Meltdown!

[airspirit]

I just lost the primary HDD on my main machine, and I'm now looking at replacing it. The thing was a piece of crap OEM Samsung 40GB drive, and I'm honestly surprised it lasted this long ....

At any rate, I'm thinking about replacing it with an 80GB Maxtor DiamondMax 7200RPM/8.5ms/8MB cache drive. I've needed to upgrade due to my huge FTP/leech collection for my network (having every version of Windows, Office, and current game available for upload takes some space), and now I'm forced to.

Problem is, is there any way to draw data off of a rapidly corrupting drive? My machine won't currently boot (it freezes when it gets to that drive), and before it did that it wouldn't boot into Windows (froze during the disk check). I'm thinking that maybe I need to re-install windows on a new disk, hook this up as a slave, and then try to leech it all to the new drive via the recovery console (bypassing Windows disk checking), but is there any damage to my mobo that can be caused by trying to run a disk that is giving up the ghost? Logically, I can't think of one (besides it refusing to boot, which isn't permanent damage), but I just want to be sure here.

I don't mind reinstalling everything, but I have a ton of stuff on there that is irreplacable and I haven't had a chance to back it all up (like saved server logs in case one of my network clients attracts the attention of the FBI). Any advice?

[cybrsamurai]

You could try running it as a slave, it might work. Is the drive making any unusual sounds or did it before it stoped functioning?
I have had several IBM drives die on me and they started making the click of doom sound before they would die completly. Even after they stopped functioning properly I was able to pull data off of them by turning the drive upside down and slaving it.

No there isnt any damage you could do to the motherboard. Its possible that you could do further damage to the data on the platters if there is a mechanical error in the HD. But if you arent going to pay someone to recover it anyway I dont think you need to worry about it.

Good Luck

[airspirit]

Well, it is fully wound all the time. It never slows ... it is like it is at full read whenever there is power to it. The noise is the same high-pitched whirring you get when you're doing a massive read.

[mfpmax]

When my 80 gig(with a nice story below) went, all I could do was put it up as slave(as mentioned) and try to get as much data off as I could...the deal is, since doing a Scandisk/Norton deal may corrupt data more(read: will) so in order to get the data off, you have to get the data off one by one and watch where the corruption starts and stops. This is kinda time consuming because if the hard drive starts having trouble accessing, you'll have to be paitent and wait for the drive to load up on screen(in my case it took 10 minutes just to view root, let alone a sub directory!).

I'd put off the DiamondMax for awhile until people stop complaining about them(me included).

So far this Segate Baracuda IV 80 gig has put up with my constant downloading and deleteion of files. The Maxtor(5400 RPM model), gave up after I had about 8 gigs total space left and started making the infamous "CLICK OF DOOM".

They'll proably sort the issue out...but right now i'd steer clear. And I like Maxtors.

[Pritorian]

I work with thoses disc inn alot of OEM computers at work and the main thing that usally breaks on the is the Electronics part.

Don´t happen to have another one lying around?

if you do its just a matter of switching the pcb and copy the things you want off it

If you don´t have another pcb, would could try putting the disk on a separat IDE channel and setting it up there. if it does not detect the disc i would think that the pcb is def.

hmmm, did that make any sence...

-P-

[TerraMex]

You really should try Pritorian's way. I've solved a couple of dead harddrives (none mine thank goodness) with that method. The bad part is... you got to have 2 pcb's .

I have a 40G Samsung Barracuda IV for a while with zero problems... but hey, i still have 2 fujitsu 10G each with no problems .

[airspirit]

Whoops ... I already bought the Maxtor drive. Oh well, it may be an excuse to buy a newer and bigger one in a couple of years (it's not like I don't do a fresh wipe/reinstall every few months anyway).

I'll give those suggestions a shot.

[jtroutma]

Airspirit:

I have been in this situation lots of times with my customers. Luckly for them, however, they did not have Samsung HDDs......

In this situation, I setup the dying HDD on its own IDE channel and setup the replacement HDD on another channel. Then I use Norton Ghost with the "ignore CRC errors" and "clone all" options checked. When it is all done cloning the HDD, you will have almost every piece of USEABLE data you can get off that HDD.

BTW If the drive has problems spinning up or reading correctly, one technique that you can try is giving the HDD a good couple "wacks" against a table. It HAS worked in the past with me but this technique offeres NO guaranties and should ONLY be used on a drive that will die anyways and you must have the data off of it.

NOTE: This is another good example why you should NEVER buy/use Samsung HDDs, ESPECIALLY in server applications; they will fail.

[airspirit]

Got most of the data off before the drive did the wind up of death. I hate hard drives sometimes.