Author Topic: Main hard drive in my PC died today suddenly (Read 31002 times)

mouser · « **on:** February 02, 2012, 03:37 AM »

This post is just a warning to all of you who are living dangerously and not performing regular backups.

This afternoon the main hard drive in my computer suddenly and unrecoverably crashed.

Luckily I had a fairly recent system backup, a spare hard drive, and a spare pc to do the image restore with.

Let this be a warning to you -- backup regularly and have a plan for recovering from a hard drive crash -- your next one could be right around the corner..

IainB · « **Reply #1 on:** February 02, 2012, 06:39 AM »

Before the disk failure, had you been running any kind of real-time disk health status/performance monitoring tool?
If so, what (if anything) did it report?

iphigenie · « **Reply #2 on:** February 02, 2012, 06:55 AM »

I have a thing for storage and backups
- use a combination of online sync/storage systems for backup - synchredible (game settings) and spideroak (others) mostly
- have 2 usb drives where (some) things are regularly sync'd and other irregularly
- have a nas where the same things are syn'd
- have genie timeline backing up selected files to either usb drive (laptop) or the NAS (desktops)

all these backups create a duplicate/version mess, but it's unlikely i'll lose a photo or document I know I want.
and yet I know that when a computer will get stolen or a drive will die, there will still be things I realise I should have backed up, or hadnt backed up a while. EIther things I forgot to add to backups, had waited to long to sync, or key files/settings I didnt realise were somewhere (eg: once made a custom object bar skin... still rue that loss as a valuable lesson)

Considered once making a site which would be a database of "whats important to back up" for all software and games... too much work but would be valuable

But it would be a hugely valuable plus for a backup software to have shareable scripts/configs that set where key data is for each software and OS.

TaoPhoenix · « **Reply #3 on:** February 02, 2012, 07:19 AM »

Heh "Too much work"

I don't have a full system image... I suppose I'll be nervous when I lose my drive - but I am trying to use my second drive as a curated save space for everything but the OS.

tomos · « **Reply #4 on:** February 02, 2012, 07:30 AM »

...but I am trying to use my second drive as a curated save space for everything but the OS.
-TaoPhoenix (February 02, 2012, 07:19 AM)

and if that one goes? (asks the guy with a chaotic backup system...)

40hz · « **Reply #5 on:** February 02, 2012, 09:20 AM »

...but I am trying to use my second drive as a curated save space for everything but the OS.
-TaoPhoenix (February 02, 2012, 07:19 AM)

and if that one goes? (asks the guy with a chaotic backup system...)
-tomos (February 02, 2012, 07:30 AM)

Yup. A quarter-hour saved now in exchange for several hours of regret sometime down the road...

With drives it isn't a question of "if." It's merely a question of when.

And Murphy is inescapable.

Stoic Joker · « **Reply #6 on:** February 02, 2012, 11:57 AM »

Despite the low-grade fiasco I went through getting a proper replacement drive. I still stand by my decision to use RAID5. Because even with the 4 array builds required before I finally ended up with the correct replacement drive ... My total down time was less than 10min. Data lost=0.

I also do regular backups of the really critical stuff to a file server, which also uses a RAID5 array.

I refuse to store a single byte in "the cloud", as I would frankly rather loose the data, than run the risk of having to explain its existence to some pompus ass in a robe...because the hosting provider got hosed in yet another bullshit sting operation.

mouser · « **Reply #7 on:** February 02, 2012, 12:30 PM »

I'm afraid i don't use any realtime drive monitoring tools.. i should.

I do have multiple drives in my pc and have a realtime backup of my documents drive to another, so i'm fairly confident there against a crash.

My backup procedures worked well this time, but there is definitely room for improvement.

Stoic Joker · « **Reply #8 on:** February 02, 2012, 02:21 PM »

I'm afraid i don't use any realtime drive monitoring tools.. i should.
-mouser (February 02, 2012, 12:30 PM)

I wasn't either, which is why my first clue was a clicking noise (oops). But I am now using the RAID controller's monitoring utility which I hadn't had installed before (because I hate clutter). Which is how I caught the subsequent (incorrect drive) failures so quickly.

IainB · « **Reply #9 on:** February 02, 2012, 06:02 PM »

...i don't use any realtime drive monitoring tools.. i should.
-mouser (February 02, 2012, 12:30 PM)

Thanks for answering my Q.
I asked because I do use such a tool, but (being paranoid), though it gets a very good write-up and has a great spec., I don't know whether it will work for sure in warning me in advance of an imminent disk failure. I therefore wanted to learn vicariously from any relevant experience you might have had. It's HDS PRO (Hard Disk Sentinel PRO) - i.e., not the freeware version - and it notified me in October 2011 that on the disk in my 11-month old laptop:

- it had logged
2011-10-11 17:26:15 #196 Reallocation Event Count 0 -> 21
2011-10-11 17:26:15 #5 Reallocated Sectors Count 0 -> 21

- that:
Performance = 100%
Health = 69% (downgraded from 100% due to the above logged events)
(This is still the current status of the disk at time of writing this post.)

And it gave me the recommendation:

There are 21 bad sectors on the disk surface. The contents of these sectors were moved to the spare area.
Based on the number of remapping operations, the health of the disk was decreased in different steps.
At this point, warranty replacement of the disk is not yet possible, only if the health drops further.
It is recommended to examine the log of the disk regularly. All new problems found will be logged there.

It is recommended to continuously monitor the hard disk status.

So, "belts-and-braces", I then ran a full system CHKDSK on the disk, and then tried to enable HDS PRO real time performance monitoring.
However, though HDS PRO can continuously monitor a disk if you tell it to, the disk has to be able to support that feature, and in my case HDS PRO said:

Real time performance monitoring is not supported on this disk.

- so I have to periodically trigger HDS PRO to recheck the disk, just in case. (So far, so good.)
Not having it automated doesn't alleviate my paranoia though!

Lesson learned:

Try to ensure that any disks bought in the future do support RT performance monitoring
- i.e., for increased peace-of-mind.

barney · « **Reply #10 on:** February 03, 2012, 04:49 PM »

It's HDS PRO (Hard Disk Sentinel PRO) - i.e., not the freeware version
-IainB (February 02, 2012, 06:02 PM)

Using the same. Also unable to run real-time checks.

However.

Yesterday, 2012-02-02, ~0200 hrs, It suddenly popped up warning me that the drive in my laptop had ~nine (9) hours to live, and I should shut down immediately and replace the drive. It also mentioned a number of sectors that had been moved, indicating surface damage.

I shut it down, switched to another box.

I rebooted the box ~six (6) hours later: no problems reported, every thing OK, remaining HD lifetime predicted to be ~1,000 days. The box I was on is a Dell Vostro with an Intel i7 CPU. It runs very hot both the CPU and the HD. And, it seems, the temperature confuses HDS Pro. I am expecting the box to die because of the heat signature, but prolly not that soon

.

Upshot of all this is that you might want to be aware that the software reports on the [purported] condition of the disk at the time it's being checked, and temperature obviously affected that report. I'm more inclined to trust trends indicated on daily reports than any single catastrophe warning. Of course, that could turn around an bite me, but it feels more comfortable and realistic, overall.

Just a bit of nosh for yer noggin.

[Edited for typos]

mouser · « **Reply #11 on:** February 03, 2012, 04:57 PM »

with the relatively low cost of hard drives i have always taken a one-strike-and-your-out replacement approach. any hint that the drive might be trouble and i backup and replace as soon as possible.

barney · « **Reply #12 on:** February 03, 2012, 08:23 PM »

Y'know, I kinda lean that way. Problem for me is the [supposed] accuracy of the reporting software. There is a drive on order, although I'll hate going through the replacement process, particularly since this'll be the first real, make-or-break test of my current backup procedure $:-\$ . On the other hand, as soon as that drive arrives, it's switch time. Then I can rehouse the suspect drive, see how it does in a standalone enclosure. But the physical aspect of replacing that drive with arthritic hands could be a major challenge

. Who knows, might even end up with spare storage

. Prolly not

.

Ath · « **Reply #13 on:** February 04, 2012, 03:48 AM »

It runs very hot both the CPU and the HD. And, it seems, the temperature confuses HDS Pro. I am expecting the box to die because of the heat signature, but prolly not that soon .
-barney (February 03, 2012, 04:49 PM)

You might wanna clean the cooling on that laptop, just to keep it alive a bit longer. Should also end the false-reporting issues with HDS.

yksyks · « **Reply #14 on:** February 04, 2012, 11:37 AM »

Regarding recovery: I had a very good experience with the DRevitalize utility, even in the demo mode (it just takes more time). Also, searching and refreshing the "slow sectors" helped me to recover several failing HDs, keeping the data intact, and they are still in a good shape. The HD must be of course recognized by the system as a disk unit, despite it might not be accessible. It worked even on some cases when the HD just made the clicking sound and nothing could be read.

Regarding full system backup: I ceased to make any more since on my two PCs the motherboard died. The full system was of obviously of no use. Backing up just the data is much easier (using Dropbox at the moment).

kyrathaba · « **Reply #15 on:** February 04, 2012, 11:41 AM »

I know you guys are gonna laugh, but I don't have that much stuff to backup: just what few C# projects I plan to continue updating, and a few other odds and ends. I just backup occasionally to memory stick and frequently to DropBox. Note that my irreplaceable photos and stuff always get backed up to both DropBox and mem-stick. And as a third measure of protection, stuff I'd really hate to lose gets FTPed to a remote repository.

mahesh2k · « **Reply #16 on:** February 04, 2012, 12:43 PM »

I do take regular backups but when processor and motherboard dies and when there is no spare computer, It is *that* uncomfortable feeling, like something is lost.

I don't take drive monitoring type of backups. I just plug USB Hard drive when my particular backup folder piles up daily.

barney · « **Reply #17 on:** February 04, 2012, 05:39 PM »

You might wanna clean the cooling on that laptop, just to keep it alive a bit longer. Should also end the false-reporting issues with HDS.
-Ath (February 04, 2012, 03:48 AM)

Pointless ... it's a Dell

, been that way since day one (1), and their supposed support contract was useless - they basically ignore[d] assistance requests. However, I did check the fan, then bought a cooling stand for it. Right now, Coretemp reports CPU @183F ... in an hour or two, HDS will report the drive to be failing, scant hours left to save data. Then I shut it down ~15 min and start all over.

OK, the point isn't the Dell, it is that HD monitoring software, particularly if it is on the drive being checked, doesn't always give a clear picture. And SMART technology only goes so far. If you're old enough, and had an electronic bent, you'll be familiar with the old tube testers ... they could not tell you a particular vacuum tube was good, only if it was definitely bad. That's sorta where SMART & HDS stand: you can get predictions, but those predictions are based upon conditions extant at the time of the report.

Like mouser, maintain a single-strike attitude. If a drive gives any indication of failing, don't rely upon it any longer than absolutely necessary. It's a mechanical device, and any mechanical device is subject to failure, even the SSDs. It pays to be paranoid about stuff like this

. Reporting software is good to have, but it can only give indications, not surety.

As far as backups go, I always remember the earlier days of MS-DOS. It was v2.3, I think. The backup [process] worked fine, but you couldn't restore it. Acronis' reputation notwithstanding (or CloneZilla's, or any other system's), I'm always dubious about the backups I make ... there've been too many restore failures

. Doesn't keep me from makin' 'em, but I don't have any real confidence in 'em

.

4wd · « **Reply #18 on:** February 04, 2012, 06:14 PM »

Regarding full system backup: I ceased to make any more since on my two PCs the motherboard died. The full system was of obviously of no use. Backing up just the data is much easier (using Dropbox at the moment).
-yksyks (February 04, 2012, 11:37 AM)

Swap the motherboard, boot into Safe Mode, uninstall old motherboard drivers, install new motherboard drivers - no different from just replacing the motherboard of a working system. I've done this a few times, usually up and running within the hour.

The only thing you need to remember is to put the SATA interfaces into IDE mode before doing anything.

A full OS image also gives you the option to run it in a VM if necessary.

Innuendo · « **Reply #19 on:** February 06, 2012, 01:25 PM »

The dirty little secret of those real-time monitoring tools is they don't monitor everything that could go wrong with a drive. There are vulnerable parts that there's just no way to monitor.

Running a monitor is better than nothing, of course, but don't think just because you do you'll always get a heads up when disaster is going to strike.

sword · « **Reply #20 on:** February 06, 2012, 02:51 PM »

My laptop hard drive became unrecoverable before New Years and I have been using Linux 'live' CDs since then. It has been great. I continue using memory sticks and transferring data to another box where I also use 'live' CDs but I installed Linux Mint v 12 to that box. Each 'live' distribution has strengths and my plan is to get an external hard drive holder (Zallman?) that I can put several .iso images on.

40hz · « **Reply #21 on:** February 06, 2012, 03:32 PM »

My laptop hard drive became unrecoverable before New Years and I have been using Linux 'live' CDs since then. It has been great. I continue using memory sticks and transferring data to another box where I also use 'live' CDs ...
-sword (February 06, 2012, 02:51 PM)

Good man!

I have a few laptops like that. I often get them from clients when the drives fail. The SATA drives are (sometimes) worth replacing depending on the laptop. But the old EIDEs aren't worth it since they're becoming harder to find - and are unacceptably expensive when you do.

But toss in a live CD and a memory stick - or plug in an external drive - and most work quite well.

These are perfect giveaways for people who just want to do some web browsing and e-mailing. I usually flip them over to senior centers or individuals who can't afford (and also don't need) much in the way of technology. Recycling is good for the soul. And good for our landfills too!

And since the OS is on read-only media, these PCs are also virtually bullet proof when it comes to malware.

Gotta love those live distro CDs.

wraith808 · « **Reply #22 on:** February 06, 2012, 04:49 PM »

So... my hard drive has been acting 'funny'. By funny, I mean it will become unrecognized from Windows (running Windows 7 32-bit) until I shut down the PC and start it up again. I've had this problem before; I thought the drive was bad so I purchased another, but changed out the SATA cable and it started working again.

Since I put the other drive that I purchased in the PC so have way too much room in it, it was trivial to copy everything from the misbehaving drive to the behaving one. Then I remembered I have HD Sentinel Pro, so started running it (duh). It tells me that everything is OK with my drive, other than one tidbit - "Hard drive failed communicating with host 4655 times". Everything is in the green... that one message is the only bad thing.

I found on hd sentinel's forums: http://www.hdsentine....php?f=18&t=1219

Thanks for the question.

Usually this problem is not related to the hard disk and I can make sure it does not cause data loss.
(if no other problem is reported, the Health should be 100%).

Such data communication problems are caused by incorrect, loose SATA cable connections.
Hard disks log all of these problems in its lifetime and this is what Hard Disk Sentinel reads and displays.

But in a new system (if proper cables and connections used), these problems can
be completely ignored as they were caused by the old system.

Now (in your system), you may even use Hard Disk Sentinel to reset the error counter:
on SMART page of the drive,
please locate the Ultra ATA CRC error count attribute.
Please locate the "Offset" column for this attribute, click on the 0 between - and +
and enter the number of problems reported but with negative sign ( -4738 ).

This will reset the error-counter to zero. So the error text will be removed from the text description
and it will be displayed again only if there will be new such problems found.

Personally I'd also recommend to use the tests in Hard Disk Sentinel to verify if the drive is
really working correctly, error-free and the stored data is safe.
Please check http://www.hdsentinel.com/faq.php#tests for more information about the
recommended steps.

I hope I could help.

Please let me know if you have any questions.
You may also use Report menu -> Send test report to developer option so I can check the
hard disk status and may advise.

But I couldn't be that unlucky with SATA cables, could I? I've checked the connections (even to the point of taking them off and putting them back on), but nothing seemed to be wrong.

*sigh* I hate hard drives.

Ath · « **Reply #23 on:** February 06, 2012, 04:56 PM »

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.

mahesh2k · « **Reply #24 on:** February 06, 2012, 04:59 PM »

Here's something that happened to me. My one laptop with linux on it died. It crashed due to overheat and now the display issue is bugging me. HP's DV2500T model has similar other problems and as per my searching on the web, it's just not me. HP support office is asking me to pay 300$ to fix display/overheating/battery, which I'm not going to do for 4 year old laptop.

I can use alternate screen to transfer data but I need S-video cable for viewing the screen. I hope linux isn't crashed inside because I want to take backup of data. It seems HP laptop was my worst buying decision.