S.M.A.R.T. reallocated sector count failure

Hello, All!

I checked the health of my drives, 2 SATA (Raid0), and 2 IDE drives. All of
them are showing ID5 reallocated sector count failures. I can understand one
showing this but all of them?? Any ideas what could cause this?

Thanks,
Colonel Blip.
E-mail:



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Colonel Blip wrote

I checked the health of my drives, 2 SATA (Raid0), and 2 IDE drives. All
of them are showing ID5 reallocated sector count failures. I can
understand one showing this but all of them?? Any ideas what could cause
this?

Some drives like some maxtors appear to deliberately ship them
that way and let the initial use sort out the marginal sectors that way.

Running the drives stinking hot can get that result too.

Presumably running the drives on a marginal power
supply can too, tho I havent actually seen that happen.

Previously Colonel Blip wrote:
Hello, All!

I checked the health of my drives, 2 SATA (Raid0), and 2 IDE drives. All of
them are showing ID5 reallocated sector count failures. I can understand one
showing this but all of them?? Any ideas what could cause this?

A bad PSU can have this effect. Overheating and mechanical shock
or vibration during operation can also cause this. There are
other potential problems that could affect all drives.

Oh, and of course the software may be misreporting things. What
are the raw reallocated count values?

Arno

Folks,

I checked with Everest and HDTune and both give the following, except that
HDTune noted all three failed the ID5.

Data - ID5 Current Worst Threshold
Data

Drive 1 227 227 63
268
Drive 2 100 100 20
1
Drive 3 (raid) 1 1 5
1883


1. Is it possible that a blue screen of death crash (hardware related) could
result in this?

2. Is it possibe the drives are all ok but had to do corrections because of
this kind of event and could be put back in order by reformatting them?

3. If 2. would work, and the backup is an image file (Ghost) would restoring
produce the same results?

Thanks,

Colonel Blip.
E-mail:

AW Previously Colonel Blip wrote:
?? Hello, All!
AW A bad PSU can have this effect. Overheating and mechanical shock
AW or vibration during operation can also cause this. There are
AW other potential problems that could affect all drives.

AW Oh, and of course the software may be misreporting things. What
AW are the raw reallocated count values?

AW Arno



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Previously Colonel Blip wrote:
Folks,

I checked with Everest and HDTune and both give the following, except that
HDTune noted all three failed the ID5.

Data - ID5 Current Worst Threshold Data
Drive 1 227 227 63 268
Drive 2 100 100 20 1
Drive 3 (raid) 1 1 5 1883

O.k., I will try an interpretation. The last column is the "raw"
value.

Drive1: 268 reallocated sectors. That is relevant. This may or may not be
a real problem. You should run a smart long self-test and see whether the
number changes. If it does the drive has a problem. If not, you
should keep an eye on it, i.e. check it every few days for some weeks.
The "normalised value" is 227, which is significantly larger than the
failed threshold of 63 (larger is better), likely out of a maximum
of 255, i.e. the attribute has dropped from 255 (perfect) to 227
(still not necessarily a problem).

Drive2: 1 reallocated sector. Not an issue at all. The drive is likely
ok. Why your tool reports a failed is beyond me. Maybe it is just
paranoid. The "normalised value" is 100, likely out of 100, i.e.
still perfect. (Some disks use 100 as "best" value, some 255, some
even mix both.)

Drive3: 1883 reallocated sectors. Bad. "Normalised value" 1, which
is below the threshold of 5. This disk actually has a failed smart
status, i.e. a value is below the threshold. This disk is dying and
it may alsready be unreadable in some areas.

Summary: Disk 1 may be o.k. or not. Disk 2 is fine. Disk 3 is dead or
dying.

Todo: - Keep an eye on the raw number of reallocated sectors of disk 1
(the last value in the attribute line) and run a long SMART
self-test on it.
- Replace Disk 3 now.

1. Is it possible that a blue screen of death crash (hardware related) could
result in this?

I doubt it.

2. Is it possibe the drives are all ok but had to do corrections because of
this kind of event and could be put back in order by reformatting them?

No. Reformatting does not work that way today. A reallocated sector is
and stays a reallocated sector. There is nothing the user can do about
it. But these are not defect secotrs. The drive already has mapped the
logical sector numbers to spare sectors. But as some time the good
spares run out and the reallocation is the sign of some more fundamental
problem, that may also kill the disk completely, possibly without
further warning.

3. If 2. would work, and the backup is an image file (Ghost) would restoring
produce the same results?

No. The sectors are allready remapped to good ones.

Arno

"Arno Wagner" wrote in message
Previously Colonel Blip wrote:
Folks,

I checked with Everest and HDTune and both give the following, except
that HDTune noted all three failed the ID5.

Data - ID5 Current Worst Threshold Data
Drive 1 227 227 63 268
Drive 2 100 100 20 1
Drive 3 (raid) 1 1 5 1883

O.k., I will try an interpretation. The last column is the "raw" value.

Drive1: 268 reallocated sectors. That is relevant. This may or may not
be a real problem. You should run a smart long self-test and see whether
the number changes. If it does the drive has a problem. If not, you
should keep an eye on it, i.e. check it every few days for some weeks.
The "normalised value" is 227, which is significantly larger than the
failed threshold of 63 (larger is better), likely out of a maximum
of 255, i.e. the attribute has dropped from 255 (perfect) to 227
(still not necessarily a problem).

268 bad sectors replaced out of millions of spares is absolute peanuts.


Drive2: 1 reallocated sector. Not an issue at all. The drive is likely
ok. Why your tool reports a failed is beyond me. Maybe it is just
paranoid. The "normalised value" is 100, likely out of 100, i.e.
still perfect. (Some disks use 100 as "best" value, some 255, some
even mix both.)

Drive3: 1883 reallocated sectors. Bad. "Normalised value" 1, which
is below the threshold of 5. This disk actually has a failed smart
status, i.e. a value is below the threshold. This disk is dying and
it may alsready be unreadable in some areas.

1883 reallocated sectors out of millions of spares is peanuts.


Summary: Disk 1 may be o.k. or not. Disk 2 is fine. Disk 3 is dead or dying.

Not with those absolute numbers. Perhaps they were accumulated during
a very short time and that that had an effect on the Normalized values
decreasing faster. If the problem stops the normalized values may return
to more normal values with time.


Todo: - Keep an eye on the raw number of reallocated sectors of disk 1
(the last value in the attribute line) and run a long SMART
self-test on it.

- Replace Disk 3 now.

It's a RAID, babble box. Which one?
The numbers are likely faked anyway.


1. Is it possible that a blue screen of death crash (hardware related) could
result in this?

I doubt it.

If the crash coincided with power failures, sure, why not.


2. Is it possibe the drives are all ok but had to do corrections because of
this kind of event and could be put back in order by reformatting them?

No. Reformatting does not work that way today.

It does on some.

A reallocated sector is and stays a reallocated sector.

Nope. Some drives can retest bad sectors and reformat tracks so that
Logical Block Addresses are sequential again.

There is nothing the user can do about it.

With some drives but not others.

But these are not defect secotrs. The drive already has mapped the
logical sector numbers to spare sectors. But as some time the good
spares run out and the reallocation is the sign of some more fundamental
problem, that may also kill the disk completely, possibly without
further warning.

Not with those absolute peanut numbers.


3. If 2. would work, and the backup is an image file (Ghost) would restoring
produce the same results?

No. The sectors are allready remapped to good ones.

But it may take care of still pending ones.


Arno

Hello, Arno!
You wrote on 24 Feb 2006 12:35:54 GMT:

Thanks for the interpretation and advice. As noted, drive 3 is a RAID array
consisting of 2 identical Maxtor drives (RAID0). Best I can tell only one
drive (1/2 of the capacity) is being reported so I'm not sure what the smart
data source is. Guess if it fails I will have to separate them and test them
as non-raid. May do that before they fail by booting on other
drive/partition.

One other question - how does on run a long SMART test?

Thanks,

Colonel Blip.
E-mail:

AW Previously Colonel Blip wrote:
?? Folks,
AW O.k., I will try an interpretation. The last column is the "raw"
AW value.



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Previously Colonel Blip wrote:
Hello, Arno!
You wrote on 24 Feb 2006 12:35:54 GMT:

Thanks for the interpretation and advice. As noted, drive 3 is a RAID array
consisting of 2 identical Maxtor drives (RAID0). Best I can tell only one
drive (1/2 of the capacity) is being reported so I'm not sure what the smart
data source is.

Good question. Depends on the controller, I guess.

Guess if it fails I will have to separate them and test them
as non-raid. May do that before they fail by booting on other
drive/partition.

You should do that as soon as possible. You can e.g. have them as
single disk on an IDE controller and use a Knoppix CD to boot. The
SMART tool there is smartctl (on the commandline).


One other question - how does on run a long SMART test?

I suse smartctl, which is a Unix tool, but also works under Windows.

For long self-test use the following commandline:

smartctl -t long device

Arno

Colonel Blip wrote

I checked with Everest and HDTune and both give the following, except
that HDTune noted all three failed the ID5.

Data - ID5 Current Worst Threshold Data

Drive 1 227 227 63 268
Drive 2 100 100 20 1
Drive 3 (raid) 1 1 5 1883

Urk, those are extremely high except for Drive 2 obviously.

1. Is it possible that a blue screen of death crash (hardware related)
could result in this?

Nope.

2. Is it possibe the drives are all ok but had to do corrections because
of this kind of event and could be put back in order by reformatting
them?

Nope.

3. If 2. would work, and the backup is an image file (Ghost) would
restoring produce the same results?

Nope, the drives have already added those bad
sectors to the bad sector list and they wont be
removed from that by restoring an image now.

What is the brand/model detail of the drives ?

Show the full SMART data for all drives, that may
have some extra useful info like the temperatures too.


Previously Colonel Blip
wrote:
Hello, All!
A bad PSU can have this effect. Overheating and mechanical shock
or vibration during operation can also cause this. There are
other potential problems that could affect all drives.

Oh, and of course the software may be misreporting things. What
are the raw reallocated count values?

Arno



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure
Usenet News==---- http://www.newsfeeds.com The #1 Newsgroup Service
in the World! 120,000+ Newsgroups ----= East and West-Coast Server
Farms - Total Privacy via Encryption =----

Folkert Rienstra wrote
Arno Wagner wrote
Colonel Blip wrote

I checked with Everest and HDTune and both give the following,
except that HDTune noted all three failed the ID5.

Data - ID5 Current Worst Threshold Data
Drive 1 227 227 63 268
Drive 2 100 100 20 1
Drive 3 (raid) 1 1 5 1883

O.k., I will try an interpretation. The last column is the "raw" value.

Drive1: 268 reallocated sectors. That is relevant. This may or may
not be a real problem. You should run a smart long self-test and see
whether the number changes. If it does the drive has a problem. If
not, you should keep an eye on it, i.e. check it every few days for some
weeks. The "normalised value" is 227, which is significantly larger
than the failed threshold of 63 (larger is better), likely out of a
maximum
of 255, i.e. the attribute has dropped from 255 (perfect) to 227
(still not necessarily a problem).

268 bad sectors replaced out of millions of spares is absolute peanuts.

Its still very high for a good drive. Something
must have produced all those bads.

Drive2: 1 reallocated sector. Not an issue at all. The drive is likely
ok. Why your tool reports a failed is beyond me. Maybe it is just
paranoid. The "normalised value" is 100, likely out of 100, i.e.
still perfect. (Some disks use 100 as "best" value, some 255, some
even mix both.)

Drive3: 1883 reallocated sectors. Bad. "Normalised value" 1, which
is below the threshold of 5. This disk actually has a failed smart
status, i.e. a value is below the threshold. This disk is dying and
it may alsready be unreadable in some areas.

1883 reallocated sectors out of millions of spares is peanuts.

Not a clue, as always.

Summary: Disk 1 may be o.k. or not.
Disk 2 is fine. Disk 3 is dead or dying.

Not with those absolute numbers. Perhaps they were accumulated
during a very short time and that that had an effect on the Normalized
values decreasing faster. If the problem stops the normalized values
may return to more normal values with time.

Not a clue, as always. It certainly indicates a problem somewhere.

Todo: - Keep an eye on the raw number of reallocated sectors of disk
1 (the last value in the attribute line) and run a long SMART
self-test on it.

- Replace Disk 3 now.

It's a RAID, babble box. Which one?
The numbers are likely faked anyway.

1. Is it possible that a blue screen of death
crash (hardware related) could result in this?

I doubt it.

If the crash coincided with power failures, sure, why not.

Very bloody unlikely to produce that many bads.

2. Is it possibe the drives are all ok but had to do
corrections because of this kind of event and could
be put back in order by reformatting them?

No. Reformatting does not work that way today.

It does on some.

**** all in fact. And those that do flout the SMART standard.

A reallocated sector is and stays a reallocated sector.

Nope. Some drives can retest bad sectors and reformat
tracks so that Logical Block Addresses are sequential again.

But they shouldnt remove reallocated sectors, stupid.

There is nothing the user can do about it.

With some drives but not others.

**** all in fact. And those that do flout the SMART standard.

But these are not defect secotrs. The drive already has mapped the
logical sector numbers to spare sectors. But as some time the good
spares run out and the reallocation is the sign of some more fundamental
problem, that may also kill the disk completely, possibly without
further warning.

Not with those absolute peanut numbers.

Not a clue, as always.

Something must have produced those very high numbers.

3. If 2. would work, and the backup is an image file
(Ghost) would restoring produce the same results?

No. The sectors are allready remapped to good ones.

But it may take care of still pending ones.

Not what he asked.