Reviving my WD 300GB Velociraptor to a usable state

[Eudorcas]

Hi,

Recently after a blackout, I started hearing strange noises coming from my HDD Velociraptor and it started freezing the windows .... Well delays has been detected with some few logical bad sectors, symptoms of a dying/failing drive; Diskeeper is my prime suspect because of its real-time defrag (I shouldn't have installed that software in the 1st place). I can't backup because BSOD welcomes me at Windows startup screen, so I necro'd it with LVL5 SpinRite and have been revived, alive & kicking for a week, then I backup everything ASAP. After a month it started freezing again and a new persistent physical bad sector appears, this one freezes even the godlike powers of SpinRite, this time I already know its dead for sure and needs replacing. Now since I still don't have the bucks to buy a new HDD, I'm running my OS in my old 320GB Barracuda IDE drive and the performance sucks.

Now my aim is to make it usable again instead of just discarding it, does have 469 realloc sect btw....
I'm just going to use it as my Installed Games storage to make my games run/load a bit faster, even if it did died and goes undetected in the BIOS, so what, I've nothing to lose. Gonna squeeze more life out of my Velociraptor before I discard it, it did cost a hefty price in the past after all.

My plan is....
1. LVL5 SpinRite - Failed, Freezing.
2. HDD Regenerator 2011 - Made it worse, hundred thousands of new bad sectors appeared after regenerating.
3. Low-Level Format, Filling the drive with Zero's - Failed because of write errors.

So my last bet is to .... Partition (isolate) the bad sectors and make good use of the good ones. No guarantee this will workout but I'll give it a try.

Anyone here have any more good ideas?

[labtech]

The internal read/write components may have weaken/damaged, which is more than just a bad sector issue.

[Eudorcas]

Already considered that, but degradation of read/write drops only when it hits the bad sectors, probably because windows switch the transfer mode from DMA 5 to PIO. It's read/write speed is still kickin in SpinRite but it freezes when it hits a bad sector and just stays there stuck up. Chances are high that the internal mechanism/components responsible for I/O are still good.

After I'm done partitioning the bad sectors, I'll do several stress testing to test its reliability.

[Doomer]

Maybe it's still under warranty?

[Eudorcas]

It's already way past its warranty. And yeah it does have 5 Year warranty but after 5 years it suddenly failed? Uh, what kind of precision timing is this?! As if there's a timed kill switch in my HDD that it'll fail after 5 years of service?
1.4 million hours MTBF (160 Years?!) and it died after 2.3 Years of service runtime (SMART Info). I better buy an UPS APC Back-Ups Pro 1500 with Ext Battery next time around.

Btw.... the isolation of physical bad sectors has been a success but I did lost almost 6 gigs of space which is not bad considering I can use it again.
It's time for some hardcore stress testing of those good sectors for reliability.

[Vulcan]

1.4 million hours MTBF (160 Years?!) and it died after 2.3 Years of service runtime (SMART Info).

By making that comparison, this shows you are misunderstanding the meaning of MTBF - there is no relationship between MTBF and service life. Research on Google (e.g. mtbf vs service life) and you will find plenty of articles. Hope those will help to explain why you should never make that comparison.

There are some other points above where I don't agree with your conclusions based on the data you provide, but since this is not a critical situation and further explanation is not important, I will just wish you good luck with the drive.

[Doomer]

5 years for a drive is very successful life
I'd say even 2 years is very successful

And I mean calendar years, not the run time years

[lcoughey]

You have too much time on your hands to make a failing drive look like it is healthy. At the end of the day, the drive is going to get worse and any future data stored on the drive will be lost.

[Eudorcas]

1.4 million hours MTBF (160 Years?!) and it died after 2.3 Years of service runtime (SMART Info).

By making that comparison, this shows you are misunderstanding the meaning of MTBF - there is no relationship between MTBF and service life. Research on Google (e.g. mtbf vs service life) and you will find plenty of articles. Hope those will help to explain why you should never make that comparison.

There are some other points above where I don't agree with your conclusions based on the data you provide, but since this is not a critical situation and further explanation is not important, I will just wish you good luck with the drive.

Already know what you mean (since 2008, learned that at TPU) .... my point is where the heck did they get that kind of ridiculously huge numbers of calculated computation (which is an Eye Candy for some people) of Mean Time Before Failures (based on predictions or intensive testing of reliability of components) since a drive can actually fail from a seconds, months to years of service life upon purchase. This MTBF is quite misleading for some people or it's been intentionally (marketing wise) put to hdd specs to mislead (layman's) people. And Thanks for the Good Luck btw.... been playing Battlefield 3 on it already I hope it'll serve me another year. Gonna populate it more with games.

You have too much time on your hands to make a failing drive look like it is healthy. At the end of the day, the drive is going to get worse and any future data stored on the drive will be lost.

The day has ended and I'm already playing Battlefield 3 on it. Not in Healthy state but pretty much in a usable state. Another year of service life is a satisfactory for me. All its going to do is read anyway.

5 years for a drive is very successful life
I'd say even 2 years is very successful

And I mean calendar years, not the run time years

It could've been longer If I did install a UPS and didn't install that dang Diskeeper with real-time defrag. With those 5 years of service it encountered countless blackouts while I'm encoding/working After Effects stuff.

Thanks for the rep guys.

[Doomer]

It could've been longer If I did install a UPS and didn't install that dang Diskeeper with real-time defrag. With those 5 years of service it encountered countless blackouts while I'm encoding/working After Effects stuff.

Don't beat yourself up
It's very unlikely a UPS would help to increase your drive's lifespan.

[BlackST]

@Doomer,

preventing "abrupt" powerdown can actually save "soft" errors, but this is not the case, it's just normal wear and tear.

I could help in Italy for refurbishing but I'm afraid we're too far (Note : sometimes there's nothing to do anyway, if it's worn out beyond repair).

Just one warning : beware of any "hdd regenerating" / "exerciser" software , they are ineffective, and also of anything that deals with "internals" (*cough!*) - you will turn your drive into a hi-tech paperweight or doorstop.



P.S. isolating a "bad zone" with HPA or partitioning won't save from normal degradation, things DO age.

[w.simon]

May be only god can bring back to life deads.

[BlackST]

It is NOT dead. It is just sick (at the moment).

It is possible to squeeze out some more life out of it but do not expect to last again more 5 years (uhm... who knows ?)

[Pixels303]

Just one warning : beware of any "hdd regenerating" / "exerciser" software , they are ineffective, and also of anything that deals with "internals" (*cough!*) - you will turn your drive into a hi-tech paperweight or doorstop.

I know, old post, but since 2007, I have been regenerating hard drives. Some drives get worse and others get better with use. Seagate being the worst of them all. Overall, I found that writing random data to sectors over 10 times actually strengthen sectors. Repeating this process every year helps keep the drive at 100%

Hard Disk Sentinel is capable of doing this to a hard drive without damaging existing data by performing a read-write new data-Write old data-Read (Verify). You can modify the options to use zeroes (default) or random data and select the number of times to perform the operation. 9/10, any weak sectors will be gone and in rare circumstances, bad sectors can be reset back to zero.

Only detail to consider, you must boot from a different volume as Hard Disk Sentinel cannot repair the boot drive. A person can clone your OS to another hard disk already in your computer and boot from that or create a custom USB boot with hard disk sentinel as a installed program, to perform the repairs.

Also, defragging the drives may trigger these bad sector events because reading from affected weak areas are what cause the bad sectors. If you do not want to add more bad sectors, make sure to backup your OS routinely and perform a multi-overwrite all sectors. This way, it strengthens all sectors without attempting to read them, preventing any additional bad sectors through CRC read errors.

HDDGURU does have a write function where you can specify a write value, and sometimes selecting patterns, 33, 55, AA, DD, FF, 00 (one pass per value for a total of 6 passes) can improve strength, but a greater number of passes with differing random values do in fact increase the strength even further.

The strengthen sector option in HDDGURU sadly does not perform as well as one would desire. My method however does decrease the CRC delays (Verified by a read hard drive sectors in HDDGURU)

I hope someone else would benefit by my experience with hard drives.



Bad sectors are usually correctable but hardware manufacturers enjoy taking away any ability to strengthening the data. Most models reserve these bad sectors permanently, so that anyone is unable to restore them back to proper use, unless you had the tools to gain direct hardware access to the drives, usually at a very high cost.