Backup Failure

seekermeister · 21 Apr 2011

I was divided on whether to post in hardware or software, but since I'm more concerned about hardware, I chose to post here.

I decided to use a backup image to create a windows 7 partition on another drive, to see if the original hard drive might be a factor in some issues that I have been trying to resolve. However, after the backup had been running for a while, it popped an error saying that a read error had occurred and that the backup had aborted.

The phraseology of the error was not clear to me as to the root cause, so I booted back into W7 and ran HD Tune on the suspect drive (screenshot below). As can be seen, it does have a bad block, but it is at nearly the end of the drive, well outside of the 100GB partition that it was working on.

Is it possible that bad block could have effected the backup? Any opinions about whether the error was the result of the software or hardware?

Mike Connor · 21 Apr 2011

It depends on what that bad block contains, and where it is located.

If you format ( Full format, not quick), the drive, then any bad blocks should be marked and ignored. You should then be able to restore an image to the drive.

If software encounters a bad block when copying larger files or restoring images etc, it often can not continue.

It is tempting to assume a linear relationship of the blocks shown in that diagram/display, to the position of blocks on the drive, but that is not really how it is. A bad block showing somewhere in a display like that merely shows the bad block in relation to how the drive is being read. The bad block could be anywhere on the drive.

Regards....Mike Connor

seekermeister · 21 Apr 2011

That is news to me, I always figured that there must be a correlation between the physical and any other manner of locating it. By what you have said, that bad block may well have been the cause of the error. I did not preformat the drive, because I have had problems with software not being able to install/run on such a partition, so I let TI format it as necessary.

I have read that it is possible to create a partition around bad sectors, so that they are fenced off, and can't be used, but if there is not a way to know precisely where that is, it would be impossible to deal with it in that manner. I guess that I will try the backup again, after having Windows format the partition, and see if TI will work with it.

There is other utilities designed to patch over bad sectors, but from what you have said, it sounds as though the partition would have to be formatted for them to work.

Wallonn7 · 21 Apr 2011

seekermeister said:

it sounds as though the partition would have to be formatted for them to work.

This does not apply if you use HDD Regenerator. If it's just "one" bad block, the trial version will solve your problem. :)

seekermeister · 21 Apr 2011

I'm probably splitting hairs, but there would be no such thing as a block on a partition that isn't formatted, would there? Since a bad block on this drive may include up to 381MBs of data, would that trial version work on that much area?

Mike Connor · 21 Apr 2011

seekermeister said:

That is news to me, I always figured that there must be a correlation between the physical and any other manner of locating it. By what you have said, that bad block may well have been the cause of the error. I did not preformat the drive, because I have had problems with software not being able to install/run on such a partition, so I let TI format it as necessary.

I have read that it is possible to create a partition around bad sectors, so that they are fenced off, and can't be used, but if there is not a way to know precisely where that is, it would be impossible to deal with it in that manner. I guess that I will try the backup again, after having Windows format the partition, and see if TI will work with it.

There is other utilities designed to patch over bad sectors, but from what you have said, it sounds as though the partition would have to be formatted for them to work.

Yes, if you have a bad block you will have to format the disk. This marks the bad block and ignores it in future. A marked/ignored block will not interfere with any operations.

The various test programs access the disks in various ways. The absolute physical location of a block on a drive does not necessarily bear any relationship to various diagrams displayed.

A drive has several actual disks, ( platters), in it. The physical location of blocks varies according to drive construction.

Some displays try to relate the physical location of the blocks as if the drive was one large disc ( Like an old style LP record), but this does not reflect reality;

http://alasir.com/books/hards/005-007.html

Regards....Mike Connor

seekermeister · 21 Apr 2011

I can understand how multiple platters would effect the interpretation of physical location, but wouldn't the displayed location, like in my screenshot, have a bearing on read location. If that is true, it would be difficult to understand how a block displayed at the utmost rear of the drive could be involved in a partition at the beginning of the drive, with many GBs between them.

Wallonn7 · 21 Apr 2011

Mike Connor said:

Yes, if you have a bad block you will have to format the disk.

At this point we disagree. My HDD Regenerator experience says otherwise.

This marks the bad block and ignores it in future. A marked/ignored block will not interfere with any operations.

But when I consider your sentence as a whole - I say this because I split it into two - we agree. :)

whs · 21 Apr 2011

I would run chkdsk /f /r to repair the block. It could have an impact if e.g. it was inside the MFT which sits up there somewhere.

Mike Connor · 21 Apr 2011

seekermeister said:

I can understand how multiple platters would effect the interpretation of physical location, but wouldn't the displayed location, like in my screenshot, have a bearing on read location. If that is true, it would be difficult to understand how a block displayed at the utmost rear of the drive could be involved in a partition at the beginning of the drive, with many GBs between them.

That might be true, but that is also why it depends on what the block contains. It also depends on how the soft/hardware reads the drive. If a bad block contains something pertaining to the location of something else, then it will cause an error when what it refers to is accessed. It could refer to anything. If it is near the perceived "end" of a disk, then such blocks often contain drive information. If you format the disk, then such a block will be marked and ignored. If it contains information then that information will normally be reconstructed. It is extremely difficult, in fact usually impossible, to determine actual physical locations of sectors on a drive, using software. What you see is an interpretation, and it does not necessarily have any bearing at all on reality.

I have often had failures due to bad blocks when trying to restore images or moving or copying some large files. Even if the blocks were apparently outside the range I was attempting to use.

Regards....Mike Connor