Any way to revive ailing drive?

[AD-15]

Hello,

I currently have a 1TB HDD for installing (almost) all my games etc onto.

Recently I've had a plethora of stability issues, crashes etc. I'd sometimes get BSODs, random restarts, freezes, the lot.

Initially, I thought it might be a result of my experimenting with enabling power saving tech whilst having an overclock set on my CPU. After a heck of a lot of tinkering, I managed to get my CPU (an i7 920) stable at 4.4GHz with EIST enabled (but not C-State tech). I thought the crashes might be a result of instability when the voltage drops down too far for the overclock in an idle state. This I can't test for using stress software, when the clock and voltage are always at the max. The way I intended to test the CPU's stability when idling on a lower voltage was just to have normal usage on my system for a few weeks, see how things went (obviously not well).

However, no matter how much I adjusted the settings in the BIOS, the crashes still occured. I figured it may be completely impossible to get power saving tech working with an OC, so I decided to give up and disabled it. Even then, the crashes continued. Also:

-The crashes weren't accompanied by the 'a clock interrupt was not received...' BSOD message, which is very typical in my experience of CPU instability

-Sometimes some functions remained for a while after a system freeze, for example a moving mouse cursor or a movie where the picture would freeze but the audio would continue (again not typical of CPU instability)

-I was often getting error messages associated with a corrupt install of Steam (which is on the HDD)

-Playing some games installed on that HDD would result in a freeze

-Copying any large amount of data from the drive would also result in a freeze

-It soon became impossible to shut down the PC, with a key clue being that the system would freeze on the shut down screen at the point where the HDD would usually spin down

-Whenever I tried to boot up I would get a CHKDSK prompt for the HDD, and the test would also soon crash the system if it wasn't skipped

I decided to try disconnecting (removing power/data cables) the drive, and this has solved all the issues. So, it looks as though there is indeed something wrong with the HDD. I'm hoping it's not a physical fault, as the drive is cool when running, there are no dodgy clicking noises etc. I'm thinking it may just be corrupted from all the system resets I had to do when messing with my CPU settings, trying to get it stable.

Is there any way I can fix the drive myself? Running the checkdisk facility in Windows 7 to try and fix any issues on the HDD causes the system to freeze.

Would appreciate any help!

[Dingo]

Unfortunately the simple answer would be to back up now while you still have the chance!!

[Thorsson]

I'd try testing the HDD in another computer, just in case it's not down to the HDD. The other thought is to boot from CD and run chkdsk from there.

[watercooled]

Does SMART tell you anything?

Also, are you using any ebay-sourced splitters/extensions? I've had a few problems with some cheap rubbish cables lately. Also try a different power connector, it's possible the current one has worn or has oxidised.

[DanceswithUnix]

I would check power to the drive and replace the SATA cable.

Personally I would use a Linux boot CD so that I could run "smartctl -a" against the drive and see what SMART was making of it.

[kompukare]

I would check power to the drive and replace the SATA cable.

Personally I would use a Linux boot CD so that I could run "smartctl -a" against the drive and see what SMART was making of it.

+1.

But I'd make my life easier by making the live Linux CD Parted Magic:

http://partedmagic.com/doku.php?id=d...s#.UQbXmESga8E

and using the GUI for smratctl (labelled 'Drive Health').

And of course, if there's any data on that drive, get that off ASAP. Sometimes Linux will mount stuff Windows can't so you never know.

[Bonebreaker777]

Try HDDScan? > HDDScan

[AD-15]

Many thanks for all the help guys.

I don't have any crucial data on the drive, so I'm not concerned about data loss. I was worried about it dying and somehow taking the other drives with it, so I've backed up any important files on there to my network drive. Even if I did need to get something off the bad drive, it would be next to impossible with any attempt at a large file transfer causing the system to crash.

Thankfully, I didn't have much to backup, as over the years I've ended up storing more and more of my important data either on the cloud or my network drive, and backing up the latter to an external drive packed safely away in a box!

I did try to reconnect my HDD today and instantly the issues returned.

All the cables I'm using are those that came standard with my PSU/Motherboard.

I tried using Parted Magic off of a live CD I made using the image linked above, but found it too difficult to work with. When I managed to get to a GUI related to SMART, any attempt to read those parameters and the program would say "drive not ready". I therefore tried using HDTune in Windows 7 to try and scan for errors. Like the basic Windows utility, as soon as it was started the system froze. I used HDDScan linked by Bonebreaker77 and it gave me the following:



I'm guessing the yellow flagged parameter isn't good?

[shaithis]

UltraDMA CRC errors sound "fairly promising".

With any luck, you will find that it's just a dodgy SATA cable.

[AD-15]

I guess I should try and swap out the SATA cable and maybe use a different SATA port too. I'm just wondering if that could confuse the file system at all when I boot into Windows?

[watercooled]

No it shouldn't matter. Also try a different SATA power connector, even on another cable string if possible.

[technodean]

I had similar issue as this with my 2tb drive before Christmas. First time it happened I took all the data off and just formatted it. few months down the line and it happened again, but in the end, Windows & Asus Bios gave me a warning that the drive was faulty and back up immediately.

[DanceswithUnix]

I guess I should try and swap out the SATA cable and maybe use a different SATA port too. I'm just wondering if that could confuse the file system at all when I boot into Windows?

Change one thing at a time else you will never truly know what the problem is.

DMA crc errors does sound like data is being corrupted between the drive and the motherboard though, so a new cable is an obvious first step.

[AD-15]

Thanks guys - will swap out the cable tonight and let you know how it goes.

[Bagnaj97]

I would check power to the drive and replace the SATA cable.

Personally I would use a Linux boot CD so that I could run "smartctl -a" against the drive and see what SMART was making of it.

smartctl -x would be better as it shows the extended info, including error logs. I've got a failing 1.5tb drive that shows the standard smart stats as ok, health status ok etc, yet it's got several errors logged which can be seen in the extended info. When I had my swap partition on the drive dmesg often had messages logged about swap errors and unreadable/unwritable sectors.

I thought the problem could be the cable, so I tried different cables/sata ports, and a different drive on the same cable and port. It's definitely the drive.

[mickyjim]

I use a program called HDAT2 it is included in Hirens boot cd and run a test i have rescued several failing drives using it