After years of not seeing any failing drives, I've had several customer PC's with them recently. I'm working on one laptop for the 2nd time, same issue: bluescreen due to bad blocks. I'm thinking its time to retire it but wonder: is there a test routine that tells one "its time" ?? If I simply look at the SMART status after doing chkdsk repairs, it will show "healthy".
??
z
I've wondered the same thing.
If you Google around, you will see a lot of recommendations to replace when bad sectors are first seen--the implication being that more bad sectors will follow, sooner rather than later.
But I've never seen any reasoning behind that assumption. Maybe more bad sectors are no more likely to occur just because you have a few. I do remember that when hard drives were very small and expensive, you didn't see the recommendation to replace offered so freely--implying that they were still serviceable.
I'm not aware of any other inexpensive consumer-level hard drive tools that shed much light on this.
Hard drives invariably have bad blocks, there is a reserve of spares, for want of a better expression at the front of the drive, the chkdsk for instance will replace and remap in the MFT, this is what is meant by repair.
Eventually there are no spares left, the disk cannot be further repaired.
Apart from the whole process inevitably slowing things down (worse than a fragged disk) it will end up with the bad blocks not being repaired.
The nearest comparison I can think of is rust on a ferrous sheet. A few spots at first but they spread eventually covering the whole surface.
That won't happen to a HDD, it will be dumped long before as it will become so corrupted it is unusable.
I have alway stood by my feelings on drives. As soon as start to wonder if it needs replacing, I that means, replace it. If you don't replace it and the client looses data, they will blame you.
One smart value of particular interest is the sector relocation count. Those are sectors that have had to be moved due to continuous read errors. A lot of drives ship with some already, but the number can go up over time. If the drive runs out of relocatable sectors, then the drive is just going to start eating data.
But like most of the smart data, a single snapshot does not usually tell you a good/bad number, you need to watch it over time, or know how certain drives behave as the numbers can mean different things on different drives
As usual, the best course of action is to have good backups and then deal with the drive failures as they happen :/ And yeah when you get a bad block bluescreen, immediate drive replacement is called for
Interesting word.
Eventually, we are all dead, along with our hard drives.
All mechanical devices are closer to failure than they were when I began this post.
How soon is the question and I don't know how you can easily find that out when it comes to hard drives. How many people even attempt to put data on a 20 or 30 year old drive?
What's the evidence that some bad blocks shown in a scan means that additional bad blocks are more likely to follow than if the scan found no bad blocks?
Maybe there is evidence--I don't know and am just asking. What would a drive engineer say?
If users with bad blocks have no backup, maybe they deserve to have a drive failure. What a concept.
I appreciate the thoughtful answers - and confirms what I was thinking. In this case, this is the second time I"ve seen this 'patient' for the same issues, and the prior was only about 3 months back. run chkdsk /r and /f - tidy it up, it will go a little longer then I bet it goes again. so I found it is a seagate model, downloaded the now-unbearably-bulky Seatools for windows [requiring .net 4] and FINALLY got that yarnball loaded, it pronounced the patient unrecoverable. fortunately, the customer has a receipt proving purchase less than 3 yrs back, so will get a replacement.
at the moment the system works fine... like nothing ever happened even though there was a good bit of file damage. Which shows my ignorance again:
When chkdsk replaces bad blocks, is there an attempt made to reconstruct the affected file or is it just moved "as is" ?
It generaly tries to save whatever bit of the file it can. In most cases the file is useless, but just in case is saves it so you can inspect it or try to recover as much data as you can.
It won't just copy "bad" data over to a good block though, it will move the file in its entirety IF the entire file can be read. Sometimes reading the same block 100 times will yield the data eventually but then the block will be marked bad. Sometimes it cant read it and in that case the file becomes a salvage attempt and it moves it but renames it like File001.chk...
makes sense.
thanx again folks
a hdd with bad blocks i would reccomend for replacment, mainly due to how people see their data as important, pictures documents and such, i would use it for programmes or games i guess after a remap done with vivard or some other programme but it would depend on just how bad it was.
i've seen drives come in with what appears to be just a few bad sectors, and be completely unusable no matter what is done, and seen drives come in that work fine all be it a little glitchy with thousands of errors.
there doesn't appear to be any consistancy with how drives react to bad sectors.
if there is important information on the drive then your best bet is to back it up and replace it asap it's not a worthwhile risk.
All modern harddisks do an automatic bad block replacement if a block is a little bit readable. I mean for example:
chkdsk bad blocks are blocks that can't be read prior to replacement described above. They are all of a sudden totally unreadable.
- First 3 attempts fail, 4th attempt succeeds. The disk drive firmware detects that problem and copies the sector to another place and gives that sector virtual the same address. The operating system doesn't even know!!! Only SMART knows. This is not seen as bad block in chkdsk
Quote