Backup Failure

[Mike Connor]

As has been stated a full format of the drive, not quick format, will mark all bad areas of the drive as not in use and will ignore, skip over, them when installing software or loading a image on the drive.

No real need for a add on program as there is no program that can make a bad block/sector/whatever good again. It is a flaw in the media.

Now the real question is will this spread? It has been my experience that once a drive starts to get bad areas it mean the drive is starting to fail.

Might be best to run the drive maker diagnostic program and see what it comes up with.

Indeed, that would be my main worry. Although it used to be quite common for new drives to have a few bad blocks, sometimes even large numbers, I have seen very little of this in recent years, and would be very wary of a drive that had them. I would not like to rely on such a drive for anything critical or important.

Regards....Mike Connor

 

[Mike Connor]

I believe that the bad sector marks, where it stores the info on what blocks are bad, are stored in a part of the disk that does not get written over by any partitioning, formatting and or file system structure.
Otherwise it would make no sense to mark these areas as bad.

That is correct, you can only reach these areas with special low-level formatting tools. Incorrect use will destroy the drive, as various essential drive information is also stored there.

Regards....Mike Connor

 

[seekermeister]

OKay, I will accept the concensus regarding location ambiguity, but I tend to doubt that methods to fix the block with something like chkdsk would work under the present circumstances, because that fix has to do with file system structure. Since TI will create a file system structure as part of it's recovery process, I would assume that it would replace any file system that was already in place, thus any fix in that file system, and TI would have the same problem with or without the fix...unless it was a permanent one, like with HDD Regenerator.

I believe that the bad sector marks, where it stores the info on what blocks are bad, are stored in a part of the disk that does not get written over by any partitioning, formatting and or file system structure.
Otherwise it would make no sense to mark these areas as bad.

You may well be correct, but from what I've read, it sounds as though both reallocated and pending sectors are recorded within the logic board, rather than the platter areas. I think that the only part of the platters that are not overwritten is the outer ring which contains the MBR and file indexes. Even the MBR normally gets overwritten, when using TI to recover an image.

 

[Mike Connor]

Just as a matter of interest, the problem with actual locations of sectors on a drive has quite serious consequences for various things, one of the major ones being the impossibility of finding a valid universal formula for optimal file placement. If you don't know where something actually is, you cant work out where it should be placed for optimal performance. This is why optimal file placement is a matter of trial and error on any particular drive,

Regards....Mike Connor

 

[Mike Connor]

OKay, I will accept the concensus regarding location ambiguity, but I tend to doubt that methods to fix the block with something like chkdsk would work under the present circumstances, because that fix has to do with file system structure. Since TI will create a file system structure as part of it's recovery process, I would assume that it would replace any file system that was already in place, thus any fix in that file system, and TI would have the same problem with or without the fix...unless it was a permanent one, like with HDD Regenerator.

I believe that the bad sector marks, where it stores the info on what blocks are bad, are stored in a part of the disk that does not get written over by any partitioning, formatting and or file system structure.
Otherwise it would make no sense to mark these areas as bad.

You may well be correct, but from what I've read, it sounds as though both reallocated and pending sectors are recorded within the logic board, rather than the platter areas. I think that the only part of the platters that are not overwritten is the outer ring which contains the MBR and file indexes. Even the MBR normally gets overwritten, when using TI to recover an image.

That is not the case, it would require some form of EEPROM on the logic board to record such things. There are no eeproms on drive logic boards. The information is contained and recorded in restricted drive sectors which can not be accessed by any normal software.

Regards....Mike Connor

 

[whs]

A quick format followed by a chkdsk is a lot faster than a full format.

 

[seekermeister]

Just as a matter of interest, the problem with actual locations of sectors on a drive has quite serious consequences for various things, one of the major ones being the impossibility of finding a valid universal formula for optimal file placement. If you don't know where something actually is, you cant work out where it should be placed for optimal performance. This is why optimal file placement is a matter of trial and error on any particular drive,

Regards....Mike Connor

How would this relate to something like O&O defrag, which has zones within which it sorts file placements? Since a defragger will move data to any number of locations, I'm unclear as to the importance of actual location, other than which partition that it is in.

 

[Mike Connor]

A quick format followed by a chkdsk is a lot faster than a full format.

True enough, and I can't think of any good reason not to do it like that. ( But I'm pretty sure there is bound to be one! )

Regards....Mike Connor

 

[seekermeister]

I believe that the bad sector marks, where it stores the info on what blocks are bad, are stored in a part of the disk that does not get written over by any partitioning, formatting and or file system structure.
Otherwise it would make no sense to mark these areas as bad.

You may well be correct, but from what I've read, it sounds as though both reallocated and pending sectors are recorded within the logic board, rather than the platter areas. I think that the only part of the platters that are not overwritten is the outer ring which contains the MBR and file indexes. Even the MBR normally gets overwritten, when using TI to recover an image.

That is not the case, it would require some form of EEPROM on the logic board to record such things. There are no eeproms on drive logic boards. The information is contained and recorded in restricted drive sectors which can not be accessed by any normal software.

Regards....Mike Connor

Again, you may be right, but this was the explanation that I read for the reason that sometimes the logic board from an identical drive will not work on another.

 

[Mike Connor]

Just as a matter of interest, the problem with actual locations of sectors on a drive has quite serious consequences for various things, one of the major ones being the impossibility of finding a valid universal formula for optimal file placement. If you don't know where something actually is, you cant work out where it should be placed for optimal performance. This is why optimal file placement is a matter of trial and error on any particular drive,

Regards....Mike Connor

How would this relate to something like O&O defrag, which has zones within which it sorts file placements? Since a defragger will move data to any number of locations, I'm unclear as to the importance of actual location, other than which partition that it is in.

All the defragmenters which run on Windows use Windows API's ( Windows Application Programming Interface). Windows API's can not access restricted sectors. For Windows, such sectors don't exist. Just as bad blocks on a formatted drive cease to exist for Windows. Some defragmenters, ( Like O&O, among others), can run an "Offline" defragmentation at boot before Windows runs, but even these can not access restricted sectors.

At one time, low level formatting software was relatively freely available, although mostly only technicians used it, but is now very difficult to obtain. You need the software for a specific drive, the manufacturers use proprietary methods of storing information.

This is still available and quite widely used;

http://hddguru.com/software/2006.04.12-HDD-Low-Level-Format-Tool/

But even this can not access from the manufacturer restricted sectors. The low-level software from the manufacturer can.

The "Optimal File Placement" strategies used by various software varies widely, and is also theoretically controversial in many respects. Windows itself "learns" where to place the files by recording how and when they are used, and places them according to usage, correct "Boot optimisation" brings significant performance increases, where simply defragmenting a drive will not.

Also many,( most), defragmentation programs do not simply defragment files, they also invariably also use whatever they think is the best file placement optimisation. In most cases this is in fact detrimental, because it gives worse results than Windows built in "learning capability".

The defragmenters have no way of knowing anything at all about your usage, or what might be an optimal file placement strategy in your case. Windows does because it analyses usage patterns. For this reason alone, ( which is seldom discussed), you are best advised to use Windows boot optimisation, and the Windows defragmenter, or at least one which recognises and does not interfere with "Layout.ini".

Regards....Mike Connor

 

[whs]

A quick format followed by a chkdsk is a lot faster than a full format.

True enough, and I can't think of any good reason not to do it like that. ( But I'm pretty sure there is bound to be one! )

Regards....Mike Connor

If you are dealing with e.g. a 1TB disk, that can be a BIG difference - especially if it is USB2 attached. if there is a good reason not to do it like that, I have not yet discovered it.

 

[Mike Connor]

A quick format followed by a chkdsk is a lot faster than a full format.

True enough, and I can't think of any good reason not to do it like that. ( But I'm pretty sure there is bound to be one! )

Regards....Mike Connor

If you are dealing with e.g. a 1TB disk, that can be a BIG difference - especially if it is USB2 attached. if there is a good reason not to do it like that, I have not yet discovered it.

I was teasing! I can't think of any reason not to do it, but I have never actually done it. There must however be a reason why a FULL format is mainly used and recommended, instead of simply doing as you suggest.

Regards....Mike Connor