hard drive woes

This can be very insiduous, as you may not realize that your data is corrupted, for a while.

i
Reply to
Ignoramus10355
Loading thread data ...

You didn't say what OS your using but if it's Windows consider using Chkdisk. You may lose some corrupt data though. I've used it and lost some corrupt data. Someone here should know if it's the best idea.

formatting link
Karl

Reply to
kfvorwerk

Whenever I've participated in a serious computer rescue effort their first step was to copy the drive.

Reply to
Jim Wilkins

I've been using CrashPlan. It looks good so far...

Reply to
David Lesher

I use rdiff-backup, personally.

i
Reply to
Ignoramus3681

But neither of these items, 2&3, are disk failures. And you have to be careful with backups. I worked my way through Engineering school as a main frame computer FE. Customer had a disk read error, so operator moved platters to another unit. No read there, and so got out backup disk, installed in on the same drive. Did not read, so moved to another drive. After getting out the oldest backup and not being able to read it, called for help. When he was done, there were no valid backups, and 56 drive heads were bad. We did not have enough spares in the San Francisco Bay area to get up all the drives. So brought up what we could and flew in more heads. Operator was looking for a new job. These were removable disk platter, 256 Mbytes drives. The drive makers should, but do not, provide disk scan software to recertify the drives, and put any bad sectors in to the drive user defect table.

Reply to
Calif Bill

Karl,

Did you resolve your issue?

Wes

Reply to
Wes

Let the Record show that "Karl Townsend" on or about Sat, 18 Apr 2009

11:23:16 -0500 did write/type or cause to appear in rec.crafts.metalworking the following:

Rescue Pro - I used it to recover a drives after the install of a DVD drive resulted in the D drive getting frotzed. Most of everything was recovered, there was some data lost. Word to the wise, the best time to install data recovery software is before the data starts to get munged.

- pyotr filipivich We will drink no whiskey before its nine. It's eight fifty eight. Close enough!

Reply to
pyotr filipivich

This is working, a little at a time. The data is movies, 4 - 9 gig per folder. I'm getting one or two before drive is warm and fails. Only 150 more trips to the freezer.

Karl

Reply to
Karl Townsend

Can you keep this drive in the freezer, but get the PC close to the freezer and connect it like that?

Or blow compressed air on the drive?

i

Reply to
Ignoramus32638

Right now I'm trying the Minnesota freezer. Took the cover off and sat it outside while still connected..

Karl

Reply to
Karl Townsend

Sounds like a great way to destroy the drive.

Reply to
Michael A. Terrell

Just curious why? Maybe you are right, but I cannot see why it would destroy the drive.

i
Reply to
Ignoramus32638

No particular reason.

I recently had a (six week old!) terabyte drive fail like this, so I attached cables, put it in a plastic bag, put it in a two quart plastic container of water and froze it into a block of ice. The drive then ran long enough to recover my files (for which I had backups, but they were very neatly arranged on this drive). And yes, I do have pictures of this setup.

I build file servers for commercial customers. I've had my fill of bad drives. I've spoken with (yelled at) sales managers and engineers from Maxtor, Seagate, Hitachi and Western Digital. I've tried to explain to them that while it's nice to be able to buy a terabyte drive for under a hundred bucks, it would be nicer to buy one for five hundred that would have a reasonable chance of surviving a few years. Their (pretty much universal) answer was, "The drive has a five year warranty. If it fails, we'll replace it." They couldn't or wouldn't understand that their drive's failure was a BIG deal to me and my customers, and that their warranty replacement involved a great deal of work and/or travel for me, and downtime for my customers. They simply don't care.

When I asked about the reliability of their enterprise drives, they answered that those carried a ten year warranty, but were not much less likely to fail. So, I burn in the drives with some torture test software for a week before I build them into a server, and I keep my fingers crossed. Fortunately the backup system in these servers is most excellent, and recovering from a failed drive is not a very big deal, but the inconvenience is horrendous.

p.

Reply to
pmv

Think about how the drive works.

Reply to
Michael A. Terrell

Hmm. After rereading this, I see the question was about compressed air, not freezing. I suppose on some drives it might be possible for compressed air to find its way in through some vent hole and disrupt the heads. Maybe.

Years ago, we were wondering about just how sensitive hard drives were to contamination. So, we opened up a drive (maybe a 40 meg Seagate) and started doing bad things to it. Fingerprints didn't do anything particularly bad. Cigarette smoke caused no errors, neither did cigarette ashes. We then poured Coca Cola right on the spinning platter, and even THAT didn't cause any errors, until the Coke dried and sort of stuck the whole thing together.

Just sayin' is all.

Reply to
pmv

I thought they the RAID systems were supposed to prevent data loss due to a drive failing. I've never tried one out but certainly I remember that was the advantage (loudly) claimed for them when they first arrived on the scene.

Cheers,

Bruce (bruceinbangkokatgmaildotcom)

Reply to
Bruce In Bangkok

No particular reason.

I recently had a (six week old!) terabyte drive fail like this, so I attached cables, put it in a plastic bag, put it in a two quart plastic container of water and froze it into a block of ice. The drive then ran long enough to recover my files (for which I had backups, but they were very neatly arranged on this drive). And yes, I do have pictures of this setup.

I build file servers for commercial customers. I've had my fill of bad drives. I've spoken with (yelled at) sales managers and engineers from Maxtor, Seagate, Hitachi and Western Digital. I've tried to explain to them that while it's nice to be able to buy a terabyte drive for under a hundred bucks, it would be nicer to buy one for five hundred that would have a reasonable chance of surviving a few years. Their (pretty much universal) answer was, "The drive has a five year warranty. If it fails, we'll replace it." They couldn't or wouldn't understand that their drive's failure was a BIG deal to me and my customers, and that their warranty replacement involved a great deal of work and/or travel for me, and downtime for my customers. They simply don't care.

When I asked about the reliability of their enterprise drives, they answered that those carried a ten year warranty, but were not much less likely to fail. So, I burn in the drives with some torture test software for a week before I build them into a server, and I keep my fingers crossed. Fortunately the backup system in these servers is most excellent, and recovering from a failed drive is not a very big deal, but the inconvenience is horrendous.

p.

Because there is not much profit in each drive as for years Compaq, Dell, etc. would get Seagate in one room, Maxtor in another, and WD in another and a couple of the other suppliers in days past and say if you give us this price you can have all our business. And then do the same thing 3 months later and buy from all the companies anyway and the disk companies always acquiesced. And then we in the disk drive world beat our suppliers to lower prices. Eventually our suppliers went out of the disk supplier business. When I was with a semiconductor supplier to the disk world, one Diskcon show they said how about building this super duper do it all chip. And our VP said put up the money and we will. But until then, we make more profit from every other part of our business than from you guys. Eventually they closed the division devoted exclusively to disk drives. Now every one wants a $50 drive.

Reply to
Calif Bill

My Dad had a TV like that once, as it warmed up the board warped and a connection would open. Dad opened a window and set the TV up so that the ass end stuck out against the storm sash. Worked fine till summer came and he didn't have time to watch TV anyhow. Gerry :-)} London, Canada

Reply to
Gerald Miller

[ ... ]

They have for me. I'm using Sun's zfs software "raidz2" with double redundancy, and so far I have not had a drive failure be unrecoverable. Pull the old drive out, (say it is connected as "c1t23d0"), plug in a known good drive to the same slot, type:

zpool replace -f bigpool c1t23d0

(where "bigpool" is the name I have given to the array made from the multiple drives), and it spends a half hour or more (depending on the drive size) "resilvering", and then returns to reporting everything as fine.

So yes, RAID (at least RAID5 and higher) can nicely survive a drive crash -- and can even keep working a bit slower with two drives down in a raidz2 configuration, and recover as the drives are replaced.

I've encountered some earlier RAID systems (at work) which tended to forget their configuration when the power went out long enough to require shutting down the computers before the UPS ran out of juice. But the zfs raidz2 (here at home) has been very forgiving -- and a good thing too, because I tend to use used disk drives picked up at hamfests and eBay auctions. As an example, one of the drives which I am currently using as part of a zfs array reports to "smartctl" the following (among many other things):

====================================================================== number of hours powered up = 57623.88 ======================================================================

So it has been around the block a few times. :-)

Enjoy, DoN.

Reply to
DoN. Nichols

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.