Serious Hardware Failure Help!!!

[Srek]

It seems that not only your PSU took damage due to the voltage spark. This is not uncommon and it's possible that the motherboard, CPU or maybe even more is fried too.
The RAID 0 configured harddisks should be readable with every other RAID controller of the same type that is configured in the same way. Using a different controller or different setup will most likely lead to loss of data (if the disk isn't fried already from the overvoltage) if you try more then simply reading data.
Good luck
Björn

[Tej]

Thanks for the reply ..... well its getting scary.... i just want to bakup my data.... the hdds were stripped in nvidia raid0 config.... on board....
so ur saying i ve to find nvidia raid controller motherboard or same mobo... then only will i b able to copy my data in case my hdd r fine? wht if they r fried... is there any way to get my data ..... its all my personal and office work on those ...... along with long list of other things.... which r really really important.....
anyways i l go check my ram on other computer.... its 400Mhz do u know if 754 pin amd will support my ram... ive got a friend with dat system... so i l go check my ram with him.... then i l find out some way to check my processor with a 939board and graphics card.... and atlast il try to change the cabinet n power supply to check the motherboard......
also when the computer crashed bak then... there were two little explosive sounds.. like thaa thaa..... one after another.... i just turned on my system then.. and just had to shut it down there and then after those sounds... the power supply was smelly so i knew it was gonner and then the ups... which is workin fine after i changed the fuse.......
but the really important thing my HDD....

well its all going to be a bit tough coz no one here is using socket 939,,,, no one is using PCIX graphics.... and i dont know if lower class cabinets with their power supplies will be able to boot my mobo.... otherwise if i cant find out...i l hav to send the mobo bak... will it be changed?? its under warranty... but they take a long time... the power supply took forever to come bak...
i guess i ve to get a new system
anyways thanks for the reply.... and tell me some way to bakup my data if u find out som other easy way...

[lots]

RAID 0 arrays are highly vunerable to data loss.

For example:
1) Should any drive in a RAID 0 array fail, ALL your data is lost.
2) Should any data on a drive in the array become corrupted, ALL your data is lost.
3) If you use a different controller, it will not understand the data on either drive.

Because of this fact, I only recommend RAID 0 to those who will ONLY be putting data there temporarily, and need it for the speed. All the data on the RAID 0 array has a back up on another drive.

Transfering your hard drives to another motherboard using the same chipset for the RAID, will work in recovering your data.

[lollygag]

I thought i was bad with my typing slang and all the ..... ... ...

ive been outdone !.. lol

[cooperunionstud]

i had the excatly the same problem when i had to rebuild my computer, but i realized that i had forgot to plug one of the power cords to the motherboard. the 4 pin one that goes next to the cpu. If you already did that then i dont really know why. but from the sound of it it looks like you might have forgot to plug one of the power cords.

[imashination]

RAID 0 arrays are highly vunerable to data loss.

For example:
1) Should any drive in a RAID 0 array fail, ALL your data is lost.
2) Should any data on a drive in the array become corrupted, ALL your data is lost.
3) If you use a different controller, it will not understand the data on either drive.


Sorry, but that set off my FUD-o-meter.

Its not highly vulnerable, in a common 2 drive system its twice as vulnerable. So if you had a 1 in a million chance before, now you have a 2 in a million; honestly nothing to worry about. If some data on the drive becomes corrupt, then only that data is lost. Suggesting any corrupt file will kill everything is just wrong

[lots]

Maybe the corruption thing was too far, but a drive failure will kill all of your data, and make it extremely difficult to recover, if at all. You can atleast recover some data (if not all) from a single drive setup.

I still wouldn't bother with a RAID 0 setup unless you have some hard drive intensive tasks (video editing). Most anything else will only see improvements in load times, which is fast enough on modern systems :P

[Tarrbot]

You know Matthew, I've said this (http://www.pcguide.com/ref/hdd/perf/raid/concepts/relRel-c.html) before. You are wrong on this but refuse to listen. RAID-0 is a very serious problem when it comes to data integrity. Believe what you will, but you can't argue with experts (http://storageadvisors.adaptec.com/2005/12/06/fata-drives/#comment-239).

From ComputerWorld (http://www.computerworld.com/blogs/node/2074):
For discussion purposes, let's consider a drive with an MTBF rating of a million hours in an enterprise class storage system with 100 drives. And let's imagine that that storage system is serving all of the mission critical applications needs of a medium sized enterprise (Email, finance, order entry and HR). To calculate the MTBF of the system we take the MTBF of the drive (1 million) and divide it by the number of drives (100 drives) to get a system MTBF of 10,000 hours. Ten thousand hours sounds like along time, but that's only 416 days or just over a year.

From MIT (http://ocw.mit.edu/NR/rdonlyres/Civil-and-Environmental-Engineering/1-264JFall-2004/EEB05C1D-A4E7-4A7A-9261-08D5E071A642/0/lectu_19_f2004_s.pdf) - Note: Link is a PDF:
Arrays with 100 disks with 500,000 hr MTBF will have failures every 5,000 hours, or every 7 months

From Hewlett Packard (http://docs.hp.com/en/146/wpunderha.pdf) - Note: Link is a PDF:
What happens when there are 200 disks on a system? Since failure rates are
multiplicative here, the MTBF for 200 disks is 200,000 hours divided by 200, or only
1000 hours! The AFR is then almost 900%, or NINE failures PER YEAR!

From Tekram (http://www.tekram.com.cn:8060/files/RAID%20Concepts%20and%20Issues.pdf): Note: Link is a PDF:
What happens if you don't include redundancy? Well, then you have a ticking time-bomb: and
that's exactly what striping without parity, RAID 0, is. A striped array without redundancy has
substantially lower reliability than a single drive and no fault tolerance. That's why I do not
recommend its use unless its performance is absolutely required, and it is supplemented with
very thorough backup procedures.

There are others from Stanford University (http://www.scs.stanford.edu/nyu/02fa/notes/l18.pdf) (PDF), University of Georgia (http://www.cs.uga.edu/~maria/classes/CSCI4730-6730/slides/RAID.ppt) (Powerpoint), and the University of Texas at Arlington (http://ranger.uta.edu/~basu/cse5350_spring06/memtech.pdf) (PDF) all saying the same exact thing.

RAID 0 is crap for data integrity.

[Tej]

Hi Guys , thanks for the suggestions.....
well i was a bit nervous with my hdds and yeh i know the raid0 will work only with the same compatible controller if u want to attach em to another motherboard.... but after tackling with everythin from changing the battery and clearing out cmos jumper,, , and still my computer didnt work so i decided to change the power supply and there it was
,,,, i was again given bak another faulty power supply,,, i wonder how cud antec guys do dat,,, i waited for 25 days or so and what again i get is another faulty power supply...
anyways i booted my computer with a scrap power supply of 300W and it worked,and i plugged in my antec 350 to another system and it didnt work although the fans of all processor,motherboard and power supply were working but it didnt boot either....
,,, so i got myself a standby power supply frm intex dat says it is 400W but i doubt it is 250 at max... coz it cost just 400Rupees(im frm india) dats like less than 10US dollars.... but my system is workin on it though....is it dangerous for my computer if i keep using this one for some time until i get another antec one??

and can we like check the actual wattage of the power supply??
and yeh i l get a big hdd like 250 or more and transfer all my important data from my raid0s to dat one...

[newman]

Antec delivering you two faulty PSUs in a row sounds a bit unbelievable. I find it much more likely the Antec's PSU has different power connections for your motherboard and you haven't plugged them all in, you should really check this out.. Make sure you plugged in all the connectors properly. Some PSUs have two power connectors for what others have one, etc. Just check the manuals and make sure you got everything right before shipping the PSU back.

edit: I forgot to put these in the post so here: ... .. ... ... ;)

[Srek]

You know Matthew, I've said this (http://www.pcguide.com/ref/hdd/perf/raid/concepts/relRel-c.html) before. You are wrong on this but refuse to listen. RAID-0 is a very serious problem when it comes to data integrity. Believe what you will, but you can't argue with experts (http://storageadvisors.adaptec.com/2005/12/06/fata-drives/#comment-239).
Easy with the statistics here please.
If you take enough highly reliable drives into account at one time chances are that one or more of them will fail almost immediately.
The larger the numbers of drives you look at is, the higher the number of drives that fail in any given period of time. However the reverse is also correct. Each individual drive is NOT very likely to fail and while two drives are worse then one it's way more often no problem then it is.
Huge numbers in statistics can "prove" anything. How many apes does it take to substitute Shakespear?

Cheers
Björn

[lots]

Well, outside of drive failure, there is alot that can go wrong with a RAID 0 array. Say your controller fails, what then? Many people tend to use on board controllers for RAID 0, which can cause problems. It seems rather silly to go buy a new motherboard if you have a controller failure, if everything else is fine. Sure the individual drive's chances of failure are generally OK, but I still would not trust RAID 0 to sensitive data, especially if it does not require the speed.

What if your power spikes, and your PSU fries and takes a few components with it (including one of your RAID'd hard drives)? Data loss is somewhat likely in that case too. RAID 0 is just more sensitive to damage than other forms of storage. You can't argue with that. I'm not giving any numbers or anything :P

Similarly, data recovery, should anything happen, is much more difficult. RAID 0 is great for the speed, if you need it. The nature of RAID 0, often prevents any success when recovering data from a failed array. If you use RAID 0, I'd suggest an equally large single drive, or RAID 1 array for main storage, while using the RAID 0 array for the work you need all that speed for. Of course you move all the sensitive data back to the storage system from your RAID 0 when you're finished.

You could also invest in a different type of RAID setup. RAID 0+1 or RAID 3 or 5, something with just a tad of redundancy :)

RAID 0 doesn't really have the right to be called RAID IMO. It lacks the redundancy, the first letter of the acronym.

[imashination]

[QUOTE=Tarrbot]You know Matthew, I've said this (http://www.pcguide.com/ref/hdd/perf/raid/concepts/relRel-c.html) before. You are wrong on this but refuse to listen. RAID-0 is a very serious problem when it comes to data integrity. Believe what you will, but you can't argue with experts (http://storageadvisors.adaptec.com/2005/12/06/fata-drives/#comment-239).

And I completely refuted and proved otherwise before, plus quoting an article where they point out that running 200 drives with no backup will lead to data loss is pointless. He wouldnt be running a 200-drive raid 0 set up. he would be running a 2-drive raid0.

The thing is im reading lots of comments like "if one drive becomes corrupt or fails, then you lose all your data!" But to point out the very obvious, if the same thing happens with a single drive (the most commen setup) then hes in exactly the same position. I seem to remember last time all sorts of crazy and incorrect statistics being thrown around where people claimed it was possible to have > 100% chance of failiure. The failure ramps up an exponential curve, have a few drives in raid0 and youre pretty much ok, once you hit a certain point then the failures will become too much.

Now regardless of all that, yes, if you can afford to buy a raid setup, then you can generally aford to buy another drive or 2 for a raid 3/5 setup. Im halfway through converting a 4 drive raid0 server at work (wtf were they thinking!?) so that everything gets backed up. lets hope it doesnt all die in the 2 weeks before im next in to finish the job eh.

[GregHess]

Just to throw in some random experience...

Having been the head administrator for the dept of entomology for the past 6 years or so (gosh has it been that long...) Harddrive failures are just part of the job. What's most interesting is the newer the harddrive, the more likely it is going to fail before 3 years are up (even when they have 5 yr warranties). It's the really old drives, namely 4-6gb 5400 rpm drives, that are still going about 7-8 yrs with no issues.

We had such a stream of failures last year (regardless of manufacturer) that I've since begun to build all new machines (or spec them) to include raid 1 arrays. I must have replaced 8 harddrives in one month last year...it was rediculous. Luckily enough people caught them early enough to allow for restorer2000pro and some crafty cloning to minimialize data loss.

And these were all machines with active cooling over the drives. Most the failures were in the 80-120 gb range. Seagate, WD, Maxtor, IBM. Mean failure rate is 2-3 years.

Thus its been my experience that it isn't "IF" a harddrive is going to fail, it's "When". Throwing 2x 250 gb's in a Raid 1 is very inexpensive and provides a level of sanity in the whole "build it faster, but keep the price down" marketing drive of the current technology line.

Just one guys opinion! You also have to remember I firmly believe there is no such thing as too much redundancy when data is involved...especially if its your bread and butter. Nobody cares if they lose a few mp3's...but 6 months of reel work? Somebody is gonna die! Standard backup at work is....
1) Raid 1 locally.
2) Raid 5 Server drives
3) External HD Backup
4) DVD Burners/Flash Drives

Extremely critical data is kept in two seperate locations! Harddrives are cheap, doctor bills aren't.

Here's a soundclip of one of the more recent failures. The harddrive actually whimpers at the end. Unfortuantly this is the sound of all your data being destroyed beyond all hope of recovery.

http://www.3dluvr.com/crossbow/incoming/harddrive.wav

[lots]

I can second Greg's experience in th IT field.

It seems like the past few months I've been calling Dell every week to get a new HD. Once, 3 times in one day.

*sigh*


As for my comment earlier about data corruption, my mistake :P i was just typing and apparently lost my mind :)

[newman]

Greg, that sound clip was... terrifying! I literally started shaking as it started the whimpering part at the end. Think I'll be having nightmares about failing HDs now... and I'm the guy who never accepted the Alien movies as Sci-Fi horrors, they were always just Sci-Fi to me.. :)

[StefanDidak]

Here's a soundclip of one of the more recent failures. The harddrive actually whimpers at the end.

Ha, "I am Raptor, hear me... whimper". I've just sent that link to Sallie (who recently presented at NYLF in San Jose in a seminar called "Backup to the Future", in which I assisted as "Vanna") because the dying HD sounds she has aren't close to being as horrific as this one. Maybe this is a good time for folks to see the Institute of Backup Trauma (http://www.backuptrauma.com/video/default2.aspx). :)

[GregHess]

Hilarous behind belief. Someone spent alot of money making that video! Thanks for sharing Stefan!

[Tarrbot]

imashination, I've been incommunicado for about 10 days and during that time I thought about what I had posted. (obviously I had not seen what you had posted at the time)

I wanted to publicly apologize to you. For some odd reason when someone says 2 in 1 million without breaking it down further, my mind reels a bit and I overreact on the numbers thing.

I've also had this same argument with noted "experts" for about 6 or 7 years who adamantly maintain that nothing is wrong as far as data "protection" is concerned with RAID 0--no matter how many drives you have.

Anyway, I just wanted to publicly apologize directly to you for coming across as an ass. I respect your opinion on here and just wanted you to know that.

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.