Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have a drive that has exactly 96 reallocated sectors for 3 years, since its purchase. It has not increased at all in number, and it has been in use constantly. I'm fairly confident it is not going to deteriorate, in that department at least, and in any case I'm willing to risk it since it holds heavily backed up data anyway.

However, the only problem is that, I assume due to to those sectors, it keeps thrashing its head back and forth for 3-4 seconds every 5-10 minutes. It does not produce any bad reading but, I assume, due to those reallocations it occasionally goes there and it results in reading the reallocation replacements from a distant part of the disk (I heard it's in the very beginning, which would explain it). This also means that it's possible it has reallocated important parts of the structure or the software run to be doing it frequently.

So, I wonder, is there a way to make that drive take those sectors that has reallocated, those 96 sectors so it never reads them, even in a logical way? I heard a simple change of partitioning would do it but it didn't fix it by simply shrinking the volume a bit in windows 8. Could it be done with a more drastic repartitioning? For example, a deletion of all partitions and a repartitioning? Though I would want to avoid that since the data is a lot.

I'd also like to avoid going directly to another drive since the laptop has no slot for a second disk so if the problem turns out to be a controller issue (the problem appears in linux and 2 versions of windows so it's most probably hardware related) I would have wasted a disk.

I'm currently thinking that, say those sectors are in the middle of the disk, physically. So, if you make one big partition is going to inevitably use them, in a logical way, and hence it might trigger that thrashing in order to use the actual good sectors in the beginning the disk in order to avoid the bad ones in the middle. Though since it does thrash relatively frequently, I wonder if they are in a relatively important or early part of the disk.

e.g. where the system holds some basic system files or where a journal is. Even in that case though I wonder if it's a simple case of just "moving" the partition to a latter part of the disk (it currently has a few other small recovery/system partitions at the back and front of it).

But in general, please try not to suggest new hardware since as I said the reallocations are stable since purchase, the data is not critical and it appears it could at least be useful as bad latency low importance storage (e.g. videos that can be deleted but it also pauses applications depended on that reading for the seconds of thrashing).

And I may buy new hardware anyway, but I still wonder if that reallocation technique is possible.

share|improve this question
If you have a Seagate/Maxtor drive have a look at the Seatools utilities - – BJ292 Jan 20 '13 at 13:59
up vote 1 down vote accepted

There is a dangerous assumption in thinking that all spare sectors are at the beginning of a disk.

They might be, but they do not need to be.

The whole 'remap a bad sector' idea is an old one. It used to be done by SCSI drives. Now I do have experience with low-level formatting SCSI drives. They used a spare sector per track and had a few emergency sectors in case all reserve sectors on the track had already been used.

Since a drive will usually read the whole track at once (even if you only ask for a single sector, just reading all the sectors when you are on the right track and buffering them will increase performance) you will not notice a slowdown when a sector gets replaced by a spare sector on the same track.

That was SCSI. SATA has borrowed many of the advanced SCSI features, and as far as I know it works precisely the same way. Which means your disk is probably not trashing due to minor reallocated sectors.

(Disclaimer: The manufacturer of the drive and its firmware really are the only ones that can confirm or deny this.)

Okay, that was assumption #1.

Now background #1:
Some drives periodically move their heads all the way across the drive to realign themselves, usually when reading fails. If this is the case, then the number of soft read errors in SMART should increase (not the reallocated number).

Background #2:
Some older drives used to periodically move their heads across the entire drive to prevent the build up of lubricant in parts of the drive. I am not sure if this is still relevant to modern drives.

Now to some answers:

1) Avoiding this by logical remapping.

Yes, if you know where the problem sectors are then you can create partitions ending before the problem place(s), and a new one starting after it.

This might be doable if they are all in one place, or near each other. If they are scattered all over the disk, then you would end up with 97 partitions. That might be a tad much. :)

2) Avoid a new drive.

Can you test the current laptop drive in another system? If it works flawlessly in there then it is definitely not a drive problem.

(No need to actually boot it. Just read the sectors you normally read a few times and see if the problem occurs. If you can not pin it down to a few sectors then read all files of the disk.)

3) Finally, the answer you want to hear:

Yes, NTFS supports a bad cluster list.

If you mark the cluster containing the remapped sector as bad then it will no longer be used by NTFS. Note that you will wipe this information if you ever reformat the drive, so write down which sectors you want to avoid.

However, I am not sure how to edit this $BadClus file.

Normally this would be done by chkdsk, but in this case the modern drive preempts chkdsk by doing its own reallocations and then not returning a bad sector.

share|improve this answer

Apologies for not directly answering your question. Feel free to ignore if this is too far off track for you.

I've found that many hard drive issues can be permanently resolved by using Spinrite. This is a fairly expensive piece of software (~US$90) but has a fantastic track record for myself and others of recovering drives that I would have otherwise written off and sorting out a load of niggling drive issues. It does a deep scan of the drive using a number of clever techniques and can sometimes completely get rid of the kind of issues you are facing.

In your case, it may be able to either fully recover the troublesome parts of the drive or permanently mark them unusable.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .