Why Doesn't SMART report this disk problem ??

[Al Dykes]

It's Fri the 13th. I rebooted a w2k machine and was suprised
to see that a NTFS disk needed a file system check and a bad block
is reported in event viewer.


I have Everest and SMART Explorer and neither report any problems
with this disk. FWIW it's a SATA disk.

WTF ?

 

[Rod Speed]

It's Fri the 13th.

You're sposed to be hiding under the
bed, not rebooting W2K machines.

I rebooted a w2k machine and was suprised
to see that a NTFS disk needed a file system
check and a bad block is reported in event viewer.

I have Everest and SMART Explorer and
neither report any problems with this disk.

If the bad doesnt show up in the Everest SMART
display, the drive cant have flagged it in the SMART
stats. Bet it does show up in the smart stats tho.

The level at which you get a popup SMART warning
error is set by whatever you are using to do that.

 

[Bob]

The level at which you get a popup SMART warning
error is set by whatever you are using to do that.

Won't the BIOS report a SMART problem if it is "enabled" in the BIOS?

My old Award BIOS does acknowledge that SMART is enabled. You would
think it would report any problems directly below that.

 

[Joep]

Won't the BIOS report a SMART problem if it is "enabled" in the BIOS?

My old Award BIOS does acknowledge that SMART is enabled. You would
think it would report any problems directly below that.

A bad block is not an error that will trigger a SMART alert. If you have
plenty of them, too many, SMART will let you know. A harddisk is equiped to
deal with a limited number of bad blocks; it will reallocate them and
replace them with spare ones.

There is a stage though when a sector can be pending reallocation, if during
this stage Windows tries to read from that block (or chkdsk), the cluster
will be flagged bad.

 

[Al Dykes]

A bad block is not an error that will trigger a SMART alert. If you have
plenty of them, too many, SMART will let you know. A harddisk is equiped to
deal with a limited number of bad blocks; it will reallocate them and
replace them with spare ones.

There is a stage though when a sector can be pending reallocation, if during
this stage Windows tries to read from that block (or chkdsk), the cluster
will be flagged bad.

I lost data and an Event ID 7 is being logged into the log.
How bad does it have to get ?

 

[Rod Speed]

Won't the BIOS report a SMART problem if it is "enabled" in the BIOS?

Not any problem. The bios will only report a level of
problems that makes it look likely that the drive is dying.

A single bad block doesnt qualify, essentially because you can get
one of those when you lose system power while the drive is writing etc.

My old Award BIOS does acknowledge that SMART is enabled.
You would think it would report any problems directly below that.

It was never intended to report any problem at all,
just give early warning that a drive may be dying.

In spades with modern drives that can and
do reallocate bad sectors automatically.

 

[Rod Speed]

I lost data and an Event ID 7 is being logged
to the log. How bad does it have to get ?

Basically bad enough to that it looks likely that the drive
is dying so you can replace it before it actually dies.

And what qualifys as bad enough to report at the bios
level is determined by the drive and thats specified by
the designers, it isnt up to the bios to decide that.

The bios is just an indication of possible imminent failure of the drive.

If you want to get notified of any errors at all, there are SMART
utes around that do report any change in the important values
like reallocated bads etc. Obviously it makes no sense to report
recovered CRC errors when all drives get those and the system
is designed to use the CRC to do retrys and get higher packing
densitys than would be possible with no errors ever seen at all.

 

[Al Dykes]

Not any problem. The bios will only report a level of
problems that makes it look likely that the drive is dying.

A single bad block doesnt qualify, essentially because you can get
one of those when you lose system power while the drive is writing etc.


It was never intended to report any problem at all,
just give early warning that a drive may be dying.

In spades with modern drives that can and
do reallocate bad sectors automatically.

Mine didn't reallocate until after the block was unrecoverable. I lost data.

Sometime soon I'll know if I can reformat the drive. In any case it's
going back to meet it's maker.

 

[Rod Speed]

Mine didn't reallocate until after the
block was unrecoverable. I lost data.

Sure, that can happen. Thats what backups are for.

Sometime soon I'll know if I can reformat the drive.

Best to try the manufacturer's diagnostic.
It should be able to map away the new bad.

 

[Al Dykes]

Sure, that can happen. Thats what backups are for.


Best to try the manufacturer's diagnostic.
It should be able to map away the new bad.

I have backups and the disk is going back tomorrow.

I want to learn why SMART didn't get me any messages
before or after I lost data.

Since the last post I've run the manufacturer's diagnostic and it says
the disk is dying and it's going back to meet it's maker and I get a
replacement.

 

[Arno Wagner]

I lost data and an Event ID 7 is being logged into the log.
How bad does it have to get ?

I had one disk that kept getting bad sectors (think it was Maxtor). I
needed to torture it some time before SMART finally showed a bad
status (which was neded to get a regular RMA over the web). I think at
that time it was up to > 1000 reallocated sectors. Some manufacturers
seem pretty optimistic on SMART failure.

Since that experience I monitor the number of reallocated sectors
directly and with the smartd from the smartmontools. The
first gives me long-term logs, the second email notification if
the value of a critical field changes.

Arno

 

[Rod Speed]

I have backups

So why are your getting your knickers in a knot ?

and the disk is going back tomorrow.

I want to learn why SMART didn't get me
any messages before or after I lost data.

I explained why and what you can do if you dont like
the level of error reporting you are currently seeing.

Since the last post I've run the manufacturer's
diagnostic and it says the disk is dying

What does the SMART data say using Everest ?

 

[Bob]

What does the SMART data say using Everest ?

As an academic exercise perhaps you can comment on the Everest report
on my boot drive.

01 Raw Read Error Rate 51 200 200 0 OK: Value is normal
03 Spin Up Time 21 104 101 3608 OK: Value is normal
04 Start/Stop Count 40 99 99 1396 OK: Value is normal
05 Reallocated Sector Count 112 199 199 2 OK: Value is normal
07 Seek Error Rate 51 200 200 0 OK: Value is normal
09 Power-On Time Count 0 53 53 34453 OK: Always passing
0A Spin Retry Count 51 100 99 1 OK: Value is normal
0B Calibration Retry Count 51 100 100 0 OK: Value is normal
0C Power Cycle Count 0 100 100 851 OK: Always passing
C4 Reallocation Event Count 0 198 198 2 OK: Always passing
C5 Current Pending Sector Count 0 200 200 0 OK: Always passing
C6 Off-Line Uncorrectable Sector Count 0 200 200 0 OK: Always
passing
C7 Ultra ATA CRC Error Rate 0 200 200 11 OK: Always passing
C8 Write Error Rate 51 200 200 0 OK: Value is normal

It would appear at first glance that this drive is still in good
condition after 5 years of use 24x7.

 

[Rod Speed]

As an academic exercise perhaps you can
comment on the Everest report on my boot drive.

Its the last value in each line thats most useful
in this situation, that column is headed Value.

01 Raw Read Error Rate 51 200 200 0 OK: Value is normal
03 Spin Up Time 21 104 101 3608 OK: Value is normal
04 Start/Stop Count 40 99 99 1396 OK: Value is normal
05 Reallocated Sector Count 112 199 199 2 OK: Value is normal

This one and its brother below are the only that indicate anything much.

07 Seek Error Rate 51 200 200 0 OK: Value is normal
09 Power-On Time Count 0 53 53 34453 OK: Always passing
0A Spin Retry Count 51 100 99 1 OK: Value is normal
0B Calibration Retry Count 51 100 100 0 OK: Value is normal
0C Power Cycle Count 0 100 100 851 OK: Always passing
C4 Reallocation Event Count 0 198 198 2 OK: Always passing
C5 Current Pending Sector Count 0 200 200 0 OK: Always passing
C6 Off-Line Uncorrectable Sector Count 0 200 200 0 OK: Always passing
C7 Ultra ATA CRC Error Rate 0 200 200 11 OK: Always passing
C8 Write Error Rate 51 200 200 0 OK: Value is normal

It would appear at first glance that this drive is
still in good condition after 5 years of use 24x7.

Yep, its fine.

 

[Bob]

This one and its brother below are the only that indicate anything much.

Yep, its fine.

That's why I am a WD fan.

I asked my friends in the IT business about WD quality and they claim
it is still as good as it was years ago. IOW, they keep buying it and
have no serious problems.

However, YMMV.

 

[Rod Speed]

That's why I am a WD fan.

No reason to be, that particular result is very very common.

And I dont care for WDs myself, too late with fluid bearings,
FAR too much desperate weaselling on warranty, too many
failures in recent times for my taste, and too noisy too.

If I was prepared to tolerate a drive noisier than the
samsungs, I might go for a seagate for the 5 year
warranty, but the Barracudas particularly get too hot
for my taste, particularly when the airflow isnt ideal.

I asked my friends in the IT business about WD quality
and they claim it is still as good as it was years ago.

They're wrong except in the sense that WD was
notorious for the 'click of death' problem at one time.

IOW, they keep buying it

More fool them.

and have no serious problems.

And plenty have had serious problems.

However, YMMV.

And then there is everyone else's too.