Main hard drive in my PC died today suddenly

[40hz]

But I couldn't be that unlucky with SATA cables, could I?

-wraith808 (February 06, 2012, 04:49 PM)

Possibly. But there could also be an electrical issue with the SATA socket on either the drive or the motherboard. Sometimes when you remove or reinsert a cable something can get minor damage which often won't manifest until the system and the drive fully heat up.

If you get the same error again try: (1) switching the cable with a known good one (2) plugging the drive into a different SATA port on the mobo. (You may need to tell your CMOS setup which port to boot off if you do. Not all mobos can autodetect the port your bootable drive is on.)

If you're still seeing errors after that, replace the drive because you've eliminated the cable and the mobo as possibles, and all that's left is the drive itself.

Don't take chances with HDs. I get more than one repeating drive error, or a non-trivial SMART alert, I'll yank them. Especially since by the time SMART detects something you're already well into the danger zone. I've had SATAs catastrophically fail with *no* SMART warnings at all.

Luck.

[tranglos]

I went through that experience less than a year ago, as chronicled here. I don't run real-time monitoring, because it does put additional strain on the system, and as my experience tells, you're likely to hear or otherwise feelthe drive failing before the software tells you:

My WD Raptor system drive died the other day after about 4 years of great performance. Lesson One: when the SMART warning kicks in, it is already too late! One moment I run the WD diagnostic tool and SMART checks out fine, a mere two hours later Windows tells me the drive has failed and needs to be replaced asap, data throughput speed drops to something like a 1.44" floppy, and you know it's going to die on you any minute.  Good thing I saw it coming hours before SMART did and made a fresh image just in time. Lesson Two: listen to your hard drive! :-)

Lesson Three: system image is a wonderful thing.

That's how it went. I noticed some operations were taking much more time than they should. After eliminating several possible causes, I realized the bottleneck was in writing data to the system drive. So I ran tests then, and they didn't detect a thing, but I knew something was very wrong. I had a recent drive image, so I backed up whatever else I could (there's no personal data on my system drive, which helps a lot), and even managed to create a full system image as well. Within two hours, the drive went into agony, and it was only then that Windows and SMART began screaming at me. Two hours too late, because at that point booting into Safe Mode took maybe 20 minutes and it was impossible to do the simplest tasks any more.

So a drive image of the system disc is a wonderful thing. Fifteen to thirty minutes and you're back in the saddle (not counting the time needed to go get a new drive). But monitoring and diagnostics utilities, not so good.

[4wd]

SATA connectors have to be one of the worst designed connectors within a computer, IMO.

A fragile friction fit connector for a critical component that's subjected to temperature cycling and vibration - just who were these brilliant engineers?

Get some decent quality cables, (FWIW, I've never had a problem with the ones that come with Gigabyte motherboards), that have the metal latching tabs.

As Ath said, throw away the old ones.

[superboyac]

eesh...yet another reminder to get my server project back into my todo list.  I have (so far) three 3TB drives sitting there, waiting to back up everything in multiply redundant fashion, as is typical of me.

I love when I tell normal people that I have 12TB of hard drives...they almost get angry with me , and their response is always along the lines of "Why do you need so much space?!"  and I'm always left thinking "Why does this make you angry??"

[wraith808]

Did you throw away the old SATA cables, and inserted new ones? That might be the only way to actually get rid of this problem.
Then use HDS Pro to reset the SMART counter.

-Ath (February 06, 2012, 04:56 PM)

I did that.  I had SATA cables from a completely different MB that I'd not had to use, and I replaced them with those.  I might actually purchase some separately to see if that helps, but I just wondered if anyone had any experience with this.

[4wd]

I might actually purchase some separately to see if that helps, but I just wondered if anyone had any experience with this.

-wraith808 (February 06, 2012, 08:21 PM)

If the SATA cable/connections are dodgy I usually get 'Delayed Write Fail' errors or an EventID of 51, 'An error was detected on device xxxxx during a paging operation.' in the Windows System Eventlog.

[wraith808]

I might actually purchase some separately to see if that helps, but I just wondered if anyone had any experience with this.

-wraith808 (February 06, 2012, 08:21 PM)

If the SATA cable/connections are dodgy I usually get 'Delayed Write Fail' errors or an EventID of 51, 'An error was detected on device xxxxx during a paging operation.' in the Windows System Eventlog.

-4wd (February 06, 2012, 08:36 PM)

Thanks for that!  I did a search of the event log, and sure enough, on the day that I was having a great many of the problems and decided to back up the HD, I was getting a bunch of errors like that.

I still don't see how I can have the same problem twice!  Guess I'm ordering more SATA cables!

[barney]

Well, so much for backups - again!

Got the new drive today.  Immediately made one (1) system image and one(1) full backup of the other partition on the old drive.  Used Acronis for the backups and to create a bootable USB stick.  Switched out the hardware.  Booted with the USB stick.  Acronis loaded.

So far, so good, right?

Then Acronis told me it had found drives - EZ-Drive, some Windows variable drive, etc., that it did not support.  Then it quit.

Went through this routine three (3) times, then the system wouldn't boot, didn't recognize anything.  Oh, yeah, on that third effort, Acronis informed me that something was 92C and it just stopped.  It didn't like the temperature so it stopped.  Didn't quit.  Didn't shut down the system.  It.just.stopped!   l let the box cool down, then tried again a couple of times - same result.

I have no faith in Acronis, despite its reputation.  For that matter, I have no faith in backups.  Why expend the effort if the damned things don't work?  And Acronis?  It should have ignored the drives it didn't like, bypassed them, then proceeded to let me at least attempt a restore.  I dislike software that is overly difficult to use.  I detest software that does not perform its advertised function, whether I paid for it (I did) or not.

So much for backups.

[Stoic Joker]

Then Acronis told me it had found drives - EZ-Drive, some Windows variable drive, etc., that it did not support.

-barney (February 08, 2012, 10:56 PM)

EZ-Drive sounds like some type of DDO (Disk Drive Overlay), and a quick google seems to agree (and produces removal instructions). I've had fits with that kind of crap in the past. My first Win2k install took months thanks to the EZ-BIOS DDO ... Until I got it torn out by the root...then all was well.

[db90h]

That's why I don't trust software backup solutions. I do a recursive file system copy with Robocopy (command line tool built into Windows) to external drives *and* a networked PC that functions solely as a backup provider for all data in this household. The one thing you don't want to do is freak out, as there is a way to restore your data (no I don't have precise instructions, but Stoic Joker seems on the right track).

Windows Backup is pretty decent these days, though still I don't trust it either. I just like a 'normal' copy of my stuff that I can easily access and retrieve without needing any supplemental software. I believe a Windows Backup Set could be recovered without Windows Backup, as it stores in common archive formats, but still I don't rely on it for critical data. I do use it though, just as another safety, in case my self-automated backups fail.

[wraith808]

Sorry about your issues with Acronis- I know they can be really terrible when it happens when you need it.  But I've *never* had it fail me in 8 years or so of using it- just figured I'd offer a contrasting opinion...

(I also use a tiered backup system... I have a copy on a NAS and online in addition to the acronis backups.  A bit paranoid... maybe... )

[db90h]

I would agree that Acronis surely works fine most all the time.

(I also use a tiered backup system... I have a copy on a NAS and online in addition to the acronis backups.  A bit paranoid... maybe... )

-wraith808 (February 09, 2012, 08:10 AM)

Not paranoid at all . You've learned (or reasoned), as I have, that backups themselves need backups, lol. The one other thing I'd like to caution people about is making sure your backups are secure, but not so secure that you can't unlock them. In other words, use a password, encrypt, but don't lose your password (or key)! It must be backed up too. Also, account for the possibility of head injury, don't rely only on your brain. As you age, you get more forgetful, and might not remember a password you set 5 years ago ;o.

[cranioscopical]

As you age, you get more forgetful, and might not remember a password you set 5 years ago

-db90h (February 09, 2012, 08:51 AM)

Or five minutes 

[tomos]

^ or move machines and realise that your backup programme - which remembers the password - is on the old (dead) one...
 

[barney]

I'm ba-a-a-ack.

Barely.

Recap:
I replaced the Dell drive, ran Acronis from a USB {stick|key|whatchamacallit}.  Acronis ran.  Kinda.  But it didn't work as an image restore.  And that's where/when I left DC.

I have a USB to IDE & SATA rig with power supply cables for various HD configurations.  I hooked the old, 'bout to fail, drive to that.  Guess what?  HDS reads that drive as perfect!  Estimated life over 1,000 days  .  This is the same drive that was in imminent fail mode, right  ?  So, now I know that HDS is subject to environmental distortion  .

I managed to pull data manually from that drive to the other laptop that is Internet enabled (my other boxes are not allowed to see the Web).  Saved all my htdocs directories.  Moved all my MySQL data.  So far, so good  .  Well, until I tried to run localhost  .  Constant connection resets, can't run any of my development stuff.  On top of that, while SQLYog can read all my MySQL files, Database Workbench 4 - in test - cannot connect, "Host hostname not allowed to connect to this MySQL server."

I've totally - ? - disabled UAC, but no juice.  Can't run local Web files, cannot reliably read databases.  All because an industry-standard imaging system wouldn't work.  (For those that know me, be prepared for a slate of In search of ...s)  Oh, yeah, I used both localhost and 127.0.0.1 to try to access htdocs, and nothing's wrong with my hosts file.

To be fair, Acronis did restore some stuff that was straight backup, not disk image.  But I can no longer trust it.  Looks as though I'll have to go back to trying to get CloneZilla to work  .

So, four (4) days totally down, another three (3) to get misc. software working.  This may not have been my ultimate week in Hell, but it ranks right up there in the top three (3)  .

A bit paranoid... maybe... )

-wraith808 (February 09, 2012, 08:10 AM)

Reminds me of a very old joke.  A king of times agone was discussing his personal protection/safety with his vizier.  When the vizier gave him that glance askance, the king said, "Oh, I know I'm paranoid.  My only concern is whether I'm paranoid enough?"

I know my paranoia just increased  .