Recovery boot sector of logical partition

[Folkert Rienstra]

Am I wrong, or the File Allocation (i.e. Order) Table of our messages
is not quite right.

Sorry, that doesn't make any sense to me.

 

[Folkert Rienstra]

Your subject line was actually correct.




Both is correct.

The drive can't read it

Or more important, reports that it can't read it ...

because the calculated ECC and the recorded ECC differ,

meaning the sector cannot be reliably read,

Or cannot be relied on ...

meaning contents is not reliable, i.e. corrupt. Also known as a bad sector.

If this state persists it can only be corrected by overwriting the sector.
In this case by using the contents of the back-up bootsector.


Not if it stays persistently bad.

Editing (from a Hex editor perspective) that is:
reading it then change it's contents and writing it back.
The reading part will not be possible.
The sector contents has to be overwritten from a different location.

 

[Eric Gisin]

No, the boot sector was written at shutdown, and the drive lost power.
I wonder how many bad sectors occur in metadata written at shutdown?

Then I must have recovered many "dead" drives. ;-) With all due respect, you
don't have the slightest idea on what you are talking about.


Would you mind explaining how come that a boot sector that was written time ago,
and was functioning properly since its creation, became all in a sudden "badly
written"?

Get this, blockhead: Boot sectors aren't rewritten occasionally and don't
become bad because they were "badly written". In many cases, such incident may
predict an imminent disk failure.

Partly correct. There is nothing to update in FAT16 and NTFS boot sectors.
There is a field in FAT32's boot sector that is updated.

I fired up diskmon and forced a dismount of my FAT32 volume. No write to 63.
I then created a test file, and NT wrote sector 63. Another write wrote 63-64.

 

[science2003]

Am I wrong, or the File Allocation (i.e. Order) Table of our messages

Sorry, that doesn't make any sense to me

Sorry for not be clear enough. Google groups beta (Italy at least)
showed on the left panel the list of messages not just in the right
order in my browser, i.e. some messages that were post presumably after
some others come up in the list before (and obviously viceversa).

 

[Folkert Rienstra]

Sorry for not be clear enough. Google groups beta (Italy at least)
showed on the left panel the list of messages not just in the right
order in my browser, i.e. some messages that were post presumably after
some others come up in the list before (and obviously viceversa).

Well, maybe that is why they call it 'beta' then.

 

[Folkert Rienstra]

No, the boot sector was written at shutdown, and the drive lost power.

Sounds like the most plausible, yes.

I wonder how many bad sectors occur in metadata written at shutdown?

One (most likely), two, ten, a hundred, does it really matter?
It must be in the tens of thousands to qualify for a 'dead as a doornail' drive.

With only one apparent bad sector it is a sure bet that overwriting that
sector will clear that same physical sector from being bad or cause a
replacement to take it's place. For that *not* to happen the drive must
either be through all of it's spares or have a dead write channel.
Either one you will be very aware of well before if that happened, unless
the latter happened at premature shutdown, that would be a real stinker.

Completely meaningless statement ....

.... and an obvious lie.

Well, as Eric just proved, you obviously do. *Not*.

Thanks Zvi, for letting me so graciously win this discussion.

They can in the death throws of a dying system when the power
fades away. Didn't they call it the Windows shutdown bug.

Not with only one sector affected. And unless this drive is killing itself
more and more by every second or minute that it is running there are
hundreds of thousands of replacement sectors to go through before
rewriting a sector won't work anymore.

Partly correct. There is nothing to update in FAT16 and NTFS boot sectors.
There is a field in FAT32's boot sector that is updated.

I fired up diskmon and forced a dismount of my FAT32 volume. No write to 63.
I then created a test file, and NT wrote sector 63. Another write wrote 63-64.

Cut the power while that happens and you have a badly written bootsector.
Now who is the blockhead here, Netiv.

 

[Zvi Netiv]

[...]
No, the boot sector was written at shutdown, and the drive lost power.

Do you know that for a fact? Where from? Or you assume that this is what
happened? You realize that the boot sector in question is of an extended
partition, not the boot one (and of a slave drive). Why would such sector be
rewritten at shutdown time? Does your assertion imply that boot sectors of all
partitions are rewritten at shutdown?

[...]

Partly correct. There is nothing to update in FAT16 and NTFS boot sectors.
There is a field in FAT32's boot sector that is updated.

What field is that (please specify its offset)?

I fired up diskmon and forced a dismount of my FAT32 volume. No write to 63.
I then created a test file, and NT wrote sector 63. Another write wrote 63-64.

Again, we are talking about sector 59,392,368, way up the drive, not the boot
partition and not even the boot drive! What is there to update on shutdown?

You should provide a better explanation than that to substantiate your
assertion.

Regards, Zvi

 

[Folkert Rienstra]

[...]
No, the boot sector was written at shutdown, and the drive lost power.

Do you know that for a fact? Where from? Or you assume that this is what
happened? You realize that the boot sector in question is of an extended
partition, not the boot one (and of a slave drive). Why would such sector be
rewritten at shutdown time? Does your assertion imply that boot sectors of all
partitions are rewritten at shutdown?

[...]

Partly correct. There is nothing to update in FAT16 and NTFS boot sectors.
There is a field in FAT32's boot sector that is updated.

What field is that (please specify its offset)?

As if it matters.

Again, we are talking about sector 59,392,368, way up the drive,
not the boot partition and not even the boot drive!

So what. He forced a dismount of his "FAT32 volume". Who said anything about
'the boot partition'. And why would a boot partition be treated any different.
You are looking for straws, Netiv.

What is there to update on shutdown?

Who cares.
Point is that it was accessed for write which makes it vulnerable when the power dies.

You should provide a better explanation than that to substantiate your assertion.

Ah, the 'it doesn't happen if you can't explain it' response.
Nice one, very convincing.