A definitive discussion about determining drive fitness

16k_zx81

Well-Known Member
Reaction score
54
Location
South Australia
I have got in a lot of drives in last couple of months with a high reallocated sector count.

If I see this I then scan the disk with software to get more information about the viability of the drive.

If this comes back 'bad' I recommend drive replacement. However this process is a bit more time-consuming than I would ideally like.

The reason for this thread is to see if its possible to drill down to a more definitive state of determining drive fitness.

What is your method for determining 'yes' or 'no' for drive viability on drives that have 'cautionary' indicators?

Jim
 
Last edited:
I have a drive on my personal laptop (I think I posted a pic of it in a thread here) that had something like 130,000 reallocated sectors. The thing ran fine and i never suspected a problem until I accidentally selected it to see the smart info. I can't speak for anyone else but usually if a drive has 20-50 reallocated sectors I consider it for replacement. Maybe it will be fine like my laptop drive but with customers machines I just don't take the chance.
 
For me, if the drive has more than 5 I will let the know the drive is failing but could last a while longer. I will often replace the drive at the clients choice or recommend they start looking for a new computer soonish. Nothing worse than seeing something, thinking "Nah, not an issue" then have it fail in a month...
 
It's really difficult to have a "line in the sand". It's more of a continuum. Generally, if the drive fails either the short or long self test because of a read error, I urge replacement asap. For reallocated or pending sectors, I suggest replacement if 15-40 and recommend replacement if 40+. I also take the drives age, noise, make and model, and format (laptop or desktop) into consideration. However, I've had drives with high number of reallocated sectors continue to run for years, as well as drives that pass all tests yet start throwing read errors two weeks later.

What I try to do is present my customer with an objective view of the situation so that they can make an informed choice.
"Based on my experience, your hard drive's condition is suspect, and I (suggest, recommend, urge) replacement. It is (possible, likely, almost certain) that your drive will soon fail completely. Should that happen,......"

I then quote a price, and we discuss. Sometimes, they are willing to take a chance with a questionable drive in an older laptop (at which point I offer our backup service). Other times, they opt to have the drive replaced.
If they are struggling to make a decision after that, I follow up with, "If it were my computer, and I used it for the same purpose as you, I would probably..(insert honest opinion here)..."

In the rare event that they decide to not replace a highly suspect drive, I document that it the "Service Notes" section on their invoice.
 
If it's the odd bad sector I don't worry about it. It's like 5 or more I tend to recommend replacement. I've no real way of knowing if this is good policy or not. I know drives can have many more bad sectors than that and run happily for years but then it's not uncommon for a few bad sectors to rapidly develop into a a lot more. Partly it depends on their backup behaviour.

As for the time consuming bit. I don't know there is a quicker way. I start at the shallow end and move up. E.g. if it fails SMART and multiple bad sectors show then I say it's bad. If it passes that I don't necessarily believe it and run the short test. If that's bad, it's bad. Then the long. This is onsite mind - where time is critical. If it's back at the workshop then these days I tend to stick UBCD in and do a long Seatools and 1/2 an hour on memtest before I do anything anyway because it saves time later when you find out you wasted 2 hours troubleshooting something with incomplete info.
 
What is your method for determining 'yes' or 'no' for drive viability on drives that have 'cautionary' indicators?

It's not for me to decide, it's up to the customer. If a drive has any cautions, I tell them the drive might last for years, or minutes...there is really no way to know with any certainty. I tell them, it's like rust on a car -- you can sand, prime and paint rust spots but new spots invariably show through in very short order. I also tell them, if they choose to keep using the drive, to be sure to back up their user data religiously, and to make a drive image to avoid an expensive re-install. If it's a good customer, I also install Foolish's Disk Health Monitor and alert them if/when I start getting alert e-mails.

There is absolutely no incentive for me to tell a customer that their drive does not need to be replaced, once Cautions start to show up.
 
"There is absolutely no incentive for me to tell a customer that their drive does not need to be replaced, once Cautions start to show up. "

^^^^^^
THIS. I always recommend replacement for any cautions detected. I explain the situation (I have a pretty good explanation I've developed for customers of what reallocated sectors are), the unpredictability of a failure, and we discuss costs.
 
Experience...seat of your pants. Combine this with age of the computer, (especially if it's after 3 years old)....if you even slightly suspect it's bad...replace it. Drives are fairly cheap. And I don't believe much in HDD diag programs..or wasting time on them. If the drive feels slow to you, had a lockup or two in its past...replace it! Done!

Clone their drive to a new one...swap out, usually done in under 20 minutes...often under 10 minutes. Client gets a much faster running computer that will have a longer life expectancy. HDDs slow down over time, as they age and get used...they get slower. New HDD..nice 'n snappy! Client is thus happier with the work you did.
 
The sound of the drives says alot to me also.

Else I run an Hitachi/IBM drive fitness test found on Hiren's.
Runs on 90% of systems.
 
Back
Top