Dell Support - how many bad blocks does it take to call a HDD dead?

[cotswoldcs]

I recently was working on a faulty Dell Vostro - 6 months old and still in warranty. The user encountered a BSOD and Windows wouldn't start, even in Safe Mode. The machine would post and start to boot and then I could hear the hard drive making an unusual noise before the BSOD. It was a Samsung 320Gb and I immediately did a test using Samsung's HUTIL utility. It reported an error - a bad block on the drive. So I phoned Dell Support to get the drive swapped out. A very nice Indian gentleman told me that a hard drive with one or two bad blocks was just fine and nothing to worry about. If the drive had more than 10 bad blocks he would be ever so happy to replace the drive for me.

10 bad blocks!!

Maybe I am being unreasonable but if the drive has one bad block then the hardware has developed a hardware fault and should be replaced. I'll admit that one bad block out of 676,000 is not many but surely there's a greater chance of data failure if more bad blocks fail. The HDD was only 6 months old and I would be concerned about using it with even 1 bad block. Or am I being unreasonable.

As it happened Dell talked me through running their own diagnostic software and in addition to the bad block it failed a SMART self short test. Only then were they happy to replace the drive.

[Jay]

1 block = faulty.

its like taking you car back with a broken front brake and them saying if the others fail then they will sort it but 3 working brakes is fine.

[gss03]

Sometimes its a bit hit and miss with Dell Techs.
Vostros unfortunately only get you through to the Indian tech support centres.

In contrast, I've had a PC disk (optiplex) called in yesterday. disk and SATA cable replaced today. Wasn't bad blocks, but it was making the PC blue screen.

[Lourdes]

Try the serial here - Samsung Warranty Service - Rexo . It'll probably come up as OEM but it's worth a try to see if you can RMA with the manufacturer.

[Splash]

1 block = faulty.

its like taking you car back with a broken front brake and them saying if the others fail then they will sort it but 3 working brakes is fine.

I've a feeling it's more like dead pixels on a TFT...

[matty-hodgson]

yeah.. but if 1 of Dells many hundreds of servers failed, that would be the 1 that had your order on it when they cock up and don't process your request for a couple of weeks!

[speardane]

soak test it (for a few days) with a low level Disk test utility - there will soon be more bad blocks!

[peterb]

All drives have bad blocks or bad sectors - and have a number (hundreds or thousands) of spares that are mapped in to replace the faulty ones as they fail over the drive's life - one of the functions of SMART monitoring is to determine how many spare sectors are left. So the user should never see bad blocks as they are hidden by the drive's internal controller. The actual condition of the drive can usually only be accessed using the mfrs connection port.

So if you are getting bad blocks reported, there is likely to be a problem with the drive and you either need to persevere with Dell as it is still in warranty, or try directly with Samsung - although you are then changing the build spec of the machine which could cause warranty problems later on if something else fails.

[IAmATeaf]

Well the chaps sort of right of sorts A drive has a number of blocks that it can mark as bad and reallocate from a pool specifically reserved for this. I've personally had and carried on using disks which have had a bad block which was then reallocated.

The main problem is, is this is an early indication that the disk is about to die and develop even more bad blocks? Nobodies can tell to be honest but I'd backup any important data just in case and I'd also make a note of where the bad block was as if 1 next to then develops a problem then it could be the signs of a creeping death.

[threepe]

Bad blocks are a really bad sign. Dell should replace the drive in my opinion, they are taking a chance with your data and time. Its not for you to soak test a new machine to see if its going to fail again, that is something that Dell should be doing.

In years gone by I have on occasion come across a HDD with a bad block, and subsequently been able to use it long term with no issues. Recently though, this has not been the case. Since hard drives have had spare blocks to assign using SMART, the user should not see bad blocks. There is a some chance that it may turn out ok in the end, but I doubt it.

Either way, you are going to have to keep very regular backups of your data.

[Lourdes]

The manufacturer would replace it so Dell should also, simple as.

[cotswoldcs]

Dell have now replaced the drive. I now have a new Western Digital 320Gb (AAKS) and all is running smoothly again.

Many thanks for your input everyone