is unhealthy drive with Powermax errors eliminated healty?

[david]

Maxtor 160GB with troubled history since I bought it in 2004 (below).
Replaced and set it aside 9 months ago.
Ran Powermax 4.23 diagnostic yesterday. Iinitially reported some
problems, no surprise.

But stops reporting problems after doing Powermax's deeper tests,
including repetitions through the night, all the tests pass now.
Surprise.

Which is stronger as an indicator of trouble-free future operation--
today's clean bill of health? or
yesterday's history of "poor health"?

In other words, does "all tests pass" mean "all problems eliminated"
or not?

Here are excerpts from the powermax log showing the sequence and
results of the tests. Once Powermax performed a write scan test, all
its tests that had failed don't fail any more.

MAXTOR POWERMAX LOG FILE
-------------------------

Installation Confirmation: PASSED

Quick Test Result: Advanced Test Required

Advanced Test Result: FAILED
Diagnostic Code: de99a87d

Burn Test Result: FAILED
Diagnostic Code: de99a86d

Quick Format Test Result: FAILED
Diagnostic Code: de99a86d

Burn Test Result: FAILED
Diagnostic Code: de99a86d

Write Scan Test Result: PASSED

Quick Test Result: PASSED

Write Scan Test Result: PASSED

Burn Test Result: PASSED

Burn Test Result: PASSED

Advanced Test Result: PASSED

Write Scan Test Result: PASSED

Write Scan Test Result: PASSED

Burn Test Result: PASSED



The history: loss of bootability several times, drive access errors in
linux, SMART error messages in linux logs, Windows disk management
utility "bad drive" message.

 

[Rod Speed]

I'd bin it if it was mine.

 

[Arno Wagner]

Maxtor 160GB with troubled history since I bought it in 2004 (below).
Replaced and set it aside 9 months ago.
Ran Powermax 4.23 diagnostic yesterday. Iinitially reported some
problems, no surprise.

But stops reporting problems after doing Powermax's deeper tests,
including repetitions through the night, all the tests pass now.
Surprise.

Which is stronger as an indicator of trouble-free future operation--
today's clean bill of health? or
yesterday's history of "poor health"?

In other words, does "all tests pass" mean "all problems eliminated"
or not?

Once bad, allways bad. With very few exceptions. If you do any
operation that overwrites the disk, bad sectors will be
reallocated. But the problem that caused the bad sectors will not
vanish, i.e. you will have more bad sectors in the future.

Arno

 

[Folkert Rienstra]

Maxtor 160GB with troubled history since I bought it in 2004 (below).
Replaced and set it aside 9 months ago.
Ran Powermax 4.23 diagnostic yesterday.
Initially reported some problems, no surprise.

But stops reporting problems after doing Powermax's deeper tests,

Which usually are destructive as opposed to the simpler read only tests.

including repetitions through the night, all the tests pass now.

Surprise.

Not really.
With nonrecoverable read error bad sectors you can run read-only tests as much as you want but nothing will change.
The drive will reassign non-recoverable read error bad sectors on writes only.

Which is stronger as an indicator of trouble-free future operation--
today's clean bill of health? or yesterday's history of "poor health"?

Depends. If the bad sectors were caused previously on a system with bad
powersupply or had suffered from temporal overheating and it is now on
a good powersupply or properly cooled it will very likely keep fine.

In other words, does "all tests pass" mean "all problems eliminated"
or not?

If they were caused externally, yes.

Here are excerpts from the powermax log showing the sequence and
results of the tests. Once Powermax performed a write scan test, all
its tests that had failed don't fail any more.

MAXTOR POWERMAX LOG FILE
-------------------------

Installation Confirmation: PASSED

Quick Test Result: Advanced Test Required

Advanced Test Result: FAILED
Diagnostic Code: de99a87d

Burn Test Result: FAILED
Diagnostic Code: de99a86d

Quick Format Test Result: FAILED
Diagnostic Code: de99a86d

Burn Test Result: FAILED
Diagnostic Code: de99a86d

Write Scan Test Result: PASSED

Quick Test Result: PASSED

Write Scan Test Result: PASSED

Burn Test Result: PASSED

Burn Test Result: PASSED

Advanced Test Result: PASSED

Write Scan Test Result: PASSED

Write Scan Test Result: PASSED

Burn Test Result: PASSED

Shows how simple bad sectors can cause an unreliable result on the diagnostic.

 

[Folkert Rienstra]

Once bad, allways bad.

Utterly clueless babblebot, as always.

With very few exceptions.

As many exceptions as opposed to being rule.

If you do any operation that overwrites the disk, bad sectors will be
reallocated.

Physical bads, yes, logical bads no.

But the problem that caused the bad sectors will not vanish,

Exactly, and if the problem is external a new drive will suffer the same faith.

i.e. you will have more bad sectors in the future.

On your new replaced drive as well.

 

[david]

Which usually are destructive as opposed to the simpler read only tests.



Not really.
With nonrecoverable read error bad sectors you can run read-only tests as much as you want but nothing will change.
The drive will reassign non-recoverable read error bad sectors on writes only.


Depends. If the bad sectors were caused previously on a system with bad
powersupply or had suffered from temporal overheating and it is now on
a good powersupply or properly cooled it will very likely keep fine.



If they were caused externally, yes.


Shows how simple bad sectors can cause an unreliable result on the diagnostic.

It's been 4 days since my original post and the drive has been under
test almost that whole 100 hours. Ran linux "shred" over it, then
"mkfs" with it's write/read of several test patterns. Then wrote my
own script to fill up the whole drive with files. After all that
re-ran Maxtor's Powermax and it passes the tests. I'm going to go
ahead and use it, just not for anything life-and-death vital.

Thanks to posters for your help.

 

[Folkert Rienstra]

It's been 4 days since my original post and the drive has been under
test almost that whole 100 hours. Ran linux "shred" over it, then
"mkfs" with it's write/read of several test patterns. Then wrote
my own script to fill up the whole drive with files. After all that
re-ran Maxtor's Powermax and it passes the tests.

I'm going to go ahead and use it, just not for anything life-and-death vital.

That's silly.
Never put "anything life-and-death vital" on any drive without having backup.

This drive of yours is not any more unreliable than any new drive that
has not yet proved itself. Any new drive can be as reliable as this one
was in it's first days of life and then go the same way this drive went.
So just treat it as you would any other drive and treat any other new
drive as you treat this one now.

 

[Odie Ferrous]

It's been 4 days since my original post and the drive has been under
test almost that whole 100 hours. Ran linux "shred" over it, then
"mkfs" with it's write/read of several test patterns. Then wrote my
own script to fill up the whole drive with files. After all that
re-ran Maxtor's Powermax and it passes the tests. I'm going to go
ahead and use it, just not for anything life-and-death vital.

I've read this thread through, and thought I might add a recent
(yesterday) experience.

I had a drive in my spares stock that was listed as having "minor
anomalies" as a result of a test run a year ago.

I needed a drive of that capacity to clone a client's hard drive to.

I ran stringent (far more so than Powermax) testing utilities on the
drive, and it tested 100%. Powermax then tested 100%.

I cloned the drive.

After I'd been working on the data for a few days, the drive packed up
on me.

I *knew* the drive was suspect, yet it still tested ok.

Following the failure yesterday, it showed up all kinds of errors;
media, heads, you name it.

Just because a drive has problems and passes a subsequent test does
*not* mean that drive is then ok.

You'd be better off binning that drive and spending the paltry amount
required for a decent drive. Something that doesn't say "Maxtor" on the
label, for starters.

If you persist on using that drive, you deserve every single problem
that will most definitely come your way.



Odie

 

[Rod Speed]

That's silly.
Never put "anything life-and-death vital" on any drive without having
backup.

This drive of yours is not any more unreliable
than any new drive that has not yet proved itself.

Mindlessly silly. We do know that that particular drive has bad
sectors. Its less clear in what situation they show up since the
SMART data doesnt keep track of the range of temperatures the
drive has seen, in spades with the voltages etc the drive has seen.

Any new drive can be as reliable as this one was in it's
first days of life and then go the same way this drive went.

Mindless waffle.

So just treat it as you would any other drive and
treat any other new drive as you treat this one now.

That would be completely stupid given that it is known
that that particular drive has seen bad sectors show up.
That is never good and it isnt clear if that is because the
drive is flakey or whether that was due to the way it was used.

 

[Arno Wagner]

I've read this thread through, and thought I might add a recent
(yesterday) experience.

I had a drive in my spares stock that was listed as having "minor
anomalies" as a result of a test run a year ago.

I needed a drive of that capacity to clone a client's hard drive to.

I ran stringent (far more so than Powermax) testing utilities on the
drive, and it tested 100%. Powermax then tested 100%.

I cloned the drive.

After I'd been working on the data for a few days, the drive packed up
on me.

I *knew* the drive was suspect, yet it still tested ok.

Following the failure yesterday, it showed up all kinds of errors;
media, heads, you name it.

Just because a drive has problems and passes a subsequent test does
*not* mean that drive is then ok.

You'd be better off binning that drive and spending the paltry amount
required for a decent drive. Something that doesn't say "Maxtor" on the
label, for starters.

If you persist on using that drive, you deserve every single problem
that will most definitely come your way.

That is consistent with my (admittedly far less extensive) experience.
If a drive has a problem serious enough to exhibit access problems,
or even trip a SMART threshold, then it has a serious defect. Serious
defects do not vanish by themselves.

The only exception I have seen so far is that a number of reallocated
defect sectors (say <500) does not necessarily signify a problem.
I have several drives (all Maxtor) with 100-300 reallocated
sectors. The numbers have been constant for almost 2 years now with
24/7 operation. I suspect these were temporary power or
vibration problems. Note that a slowly (or fast) increasing number
of reallocated secors does indicate a persistent problem and
should be taken very seriously.

Arno

 

[david]

The debate on this occult science is very interesting. For what it's
worth, I did go out and buy another 160GB drive, new, yesterday. (Says
"Western Digital")

But at the same time, I'm going to give the old Maxtor continued use.
I'm not sure what I'll use it for, doesn't matter, I'm interested in
seeing how long or short its performance life is going to be. Pure
curiosity is aroused now. I'll be careful which drive I put "real"
data on, and keep that data backed up in any case.

Thanks for posts.

 

[Folkert Rienstra]

Good to know that you keep faulty drives for your clients to do your
recoveries with.

So it was 100%.

Gosh, you killed it.

Thanks for so obviously making that up, Odoriferous.

Except -of course- if it is a new out of the box one, right.

That is consistent with my (admittedly far less extensive) experience.

Nonetheless, even you with your "admittedly far less extensive experience"
experienced the deviation from that consistency.

If a drive has a problem serious enough to exhibit access problems,
or even trip a SMART threshold, then it has a serious defect.

Nonsense. You can write the drive full of bad sectors, read them a few
hundred times and trip a treshold. There's nothing wrong with it because
it was YOU who wrote them, not any serious condition with the drive.

Serious defects do not vanish by themselves.

Physical ones, yes. Logical ones do vanish, once overwritten.

The only exception I have seen so far is that a number of reallocated
defect sectors (say <500) does not necessarily signify a problem.

I have several drives

Several no less. Rather odd definition of "only" you have.
If you have "only" several as an exception, you must have hundreds die of
other causes. Some might say -not me of course- that you are killing them.

(all Maxtor) with 100-300 reallocated sectors. The numbers have been
constant for almost 2 years now with > 24/7 operation. I suspect these
were temporary power or vibration problems.

Stupid babblebot disagrees with me and then goes on to confirm what I say.

Note that a slowly (or fast) increasing number of reallocated secors
does indicate a persistent problem and should be taken very seriously.

Thanks, babblebot, for putting AND Odoriferous AND yourself to shame.

 

[Odie Ferrous]

Good to know that you keep faulty drives for your clients to do your
recoveries with.

Actually, that was, as I said, in my "spares" stock - not my "float"
stock.

"Spares" stock consists of drives that are suspect, but which have at
least one serviceable component to use for recovery work.

"Float" stock consists of drives that I purchased new.

On this occasion, I wanted an 80GB drive to take a clone from another
80GB drive. For *temporary* file sorting and manipulation, after which
it would have been copied to a main recovery machine for a more secure
backup.

Instead of spending your life nit-picking, why not consider taking on
some charity work to keep your mind occupied?



Odie

 

[Rod Speed]

Folkert Rienstra wrote

Actually, that was, as I said, in my "spares" stock - not my "float" stock.

"Spares" stock consists of drives that are suspect, but which have
at least one serviceable component to use for recovery work.

"Float" stock consists of drives that I purchased new.

On this occasion, I wanted an 80GB drive to take a clone from another
80GB drive. For *temporary* file sorting and manipulation, after
which it would have been copied to a main recovery machine for a more
secure backup.

Instead of spending your life nit-picking,

That aint anything even remotely resembling anything like a 'life'

why not consider taking on some charity work to keep your mind occupied?

Wouldnt work. Those rabid bloodshot eyes and the flecks of
foam about the lips keep putting everyone off for some reason.

 

[Folkert Rienstra]

Right, and you used one of these suspect drives as an intermediate to " a more
secure backup". Does the term "the chain is as strong as the weakest shackle"
ring a bell, Odoriferous? Nice to know what you consider "secure", thanks.

Right, which must have been empty then (as in non-existent?) if you did all
this effort to get a suspect drive working. Time is money right, and with
80-hour workweeks you have so little of it, especially if you have to also
troll -not one, but several- news groups in the little time left after that.

That aint anything even remotely resembling anything like a 'life'

Which you know of first hand, isn't it, Roddles.

Wouldnt work. Those rabid bloodshot eyes and the flecks of
foam about the lips keep putting everyone off for some reason.

Which your fanclubs are the living proof of.