The life of a hard drive?

How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?

--
"All men make mistakes, but married men find out about them sooner."
-- Red Skelton

"Ablang" wrote in message
...
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?


As for many products the failure rate is 'bath' shaped, high at the start
(like yours
then very low for a long time untill ageing starts to take effect.
If it lasts a year it is lilkely last a lifetime these days, a lifetime
being so long that
effectively it would be obsolete by the time it failed, an example would be
a
20 megabyte (not giga) drive, many of which may still work today but no one
would every use one today.

--
"All men make mistakes, but married men find out about them sooner."
-- Red Skelton

In article ,
half_pint wrote:

"Ablang" wrote in message
...
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?


As for many products the failure rate is 'bath' shaped, high at the start
(like yours
then very low for a long time untill ageing starts to take effect.
If it lasts a year it is lilkely last a lifetime these days, a lifetime
being so long that
effectively it would be obsolete by the time it failed, an example would be
a
20 megabyte (not giga) drive, many of which may still work today but no one
would every use one today.

The formal definition of the the MTBF rating is the # of hours over
which 50% of the drives would fail, BUT, only applies within the
stated lifetime of the product. So, given a 30 year MTBF and a 5year
life specification one twelveth of the drives will fail. I think I've
got that right. After 5 years all bets are off.

You can pull the spec sheet for any drive model from the manufatcurers
web site. Look for lifetime, MTBF, and recommended max operating
temperature. Keep yoiu rdrive cool.





--
Al Dykes
-----------
adykes at p a n i x . c o m

On Wed, 14 Apr 2004 01:20:04 +0000 (UTC), Ablang
wrote:

How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?

I tend not to keep them that long but usually end up selling them
cheap or giving them away to someone I know so Ive seen them last a
long time. The thing is people I know tend not to use them that hard
and I tend to work their systems so they are put in a super small case
with no ventilation crammed between two other devices.

They tend to last a long time. Lets see - most people I know have
upgraded but still use a HD that they bought 4-7 yrs ago as a second
or third HD usually around 6-20 gigs.

Im working on a neighbors system now that has two HDs 20 gig that are
dated at 1999. Tell you the truth Ive NEVER had a HD fail on my
personal sysem and only one failed in a system I worked on.

The one that did - I bought 6 yrs ago when HDs were relatively still
expensive. I saw a local mom and pop selling what was usually cheap at
the time - some leftover seagates that were of a dated design. I think
they were actually Conners - didnt Seagate or quantum buy Conner?
Anyway they werent the seagates in the current line up at the time.
They were kind of suspect HDs - dated and I couldnt even find the
actual model specs. I put it in a friends system who hardly used their
PC at all. The thing failed after 1.5 yrs and Im not sure what it went
through since they moved twice in that time. Was it dropped ? Who
knows.

Seagate sent a replacement though and thats lasted for 4 yrs after
that. I still have a Maxtor thats 120 MEGs !!!! It still runs. I think
I had it on a 486 system ages ago . Cant even remember when I bought
it.

On Wed, 14 Apr 2004 02:42:14 +0100, "half_pint"
wrote:


"Ablang" wrote in message
...
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?


As for many products the failure rate is 'bath' shaped, high at the start
(like yours
then very low for a long time untill ageing starts to take effect.
If it lasts a year it is lilkely last a lifetime these days, a lifetime
being so long that
effectively it would be obsolete by the time it failed, an example would be
a
20 megabyte (not giga) drive, many of which may still work today but no one
would every use one today.

I don't know about that "lasts a year it'll last a lifetime" part... to
me it seems more accurate that if it lasts 3 months it'll likely last 5
years but after that all bets are off, it becomes too much of a liability
to continue using it, and even shorter replacement interval is warranted
with more valuable data, not even considering the performance or capacity
benefits.

kony wrote:
On Wed, 14 Apr 2004 02:42:14 +0100, "half_pint"
wrote:


"Ablang" wrote in message
...
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years,
but I had a WD 6.x GB HD that just became as good as dead a few
weeks ago (too many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?


As for many products the failure rate is 'bath' shaped, high at the
start (like yours
then very low for a long time untill ageing starts to take effect.
If it lasts a year it is lilkely last a lifetime these days, a
lifetime being so long that
effectively it would be obsolete by the time it failed, an example
would be a
20 megabyte (not giga) drive, many of which may still work today but
no one would every use one today.

I don't know about that "lasts a year it'll last a lifetime" part...
to me it seems more accurate that if it lasts 3 months it'll likely
last 5 years but after that all bets are off, it becomes too much of
a liability to continue using it, and even shorter replacement
interval is warranted with more valuable data, not even considering
the performance or capacity benefits.

After many years in IT support I'd have to say the main reasons for drives
ending their life are in order:

Becoming Obsolete
Physical Abuse
Component Failure

With the order reversed during the first few months.

Alan

Ablang wrote:
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?

IDE drives are NOT intended to run over 50% service (on for 12 hours,
OFF for 12 hours!)

My best drives are SCSI, and the 10,000rpm Cheetah by Seagate last 7
years in 100% service, on servers! (my records). Zero failurs in 7 years.

Western Digital, Maxtor, Fujitsu, Hitachi, Seagate, and IBM IDE drives
seem to do OK on my 100% systems, but, I see from my records, that they
get to be RMA under warranty about 30%. So, about one out of three IDE
drives fails every two years!

The failed drives start doing a slow click, click click...
That lets me know that it is time to backup, R&R...
I run SCSI on these systems, also, so I jsut move the stuff over to a
SCSI drive!

"Alan Walker" wrote in message
...
kony wrote:
On Wed, 14 Apr 2004 02:42:14 +0100, "half_pint"
wrote:


"Ablang" wrote in message
...
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years,
but I had a WD 6.x GB HD that just became as good as dead a few
weeks ago (too many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?


As for many products the failure rate is 'bath' shaped, high at the
start (like yours
then very low for a long time untill ageing starts to take effect.
If it lasts a year it is lilkely last a lifetime these days, a
lifetime being so long that
effectively it would be obsolete by the time it failed, an example
would be a
20 megabyte (not giga) drive, many of which may still work today but
no one would every use one today.

I don't know about that "lasts a year it'll last a lifetime" part...
to me it seems more accurate that if it lasts 3 months it'll likely
last 5 years but after that all bets are off, it becomes too much of
a liability to continue using it, and even shorter replacement
interval is warranted with more valuable data, not even considering
the performance or capacity benefits.

After many years in IT support I'd have to say the main reasons for drives
ending their life are in order:

Becoming Obsolete
Physical Abuse
Component Failure

With the order reversed during the first few months.

After many years of using disks, all the way back to the old Bryant 2A with
3.5 FOOT platters (honest!), I believe that today, with the modern
technology, dropping a disk is the biggest cause of failure followed by
bearing wear. I have never had a disk fail that I could prove was due to
positioner problems or to the electronics failing. Cables can be a problem.
Power supplies don't seem to be a cause either.

On Wed, 14 Apr 2004 13:26:11 GMT, Patrick wrote:

IDE drives are NOT intended to run over 50% service (on for 12 hours,
OFF for 12 hours!)

Sources please? I've never heard this.

Western Digital, Maxtor, Fujitsu, Hitachi, Seagate, and IBM IDE drives
seem to do OK on my 100% systems, but, I see from my records, that they
get to be RMA under warranty about 30%. So, about one out of three IDE
drives fails every two years!

I've got a lab full of IDE systems running 24/7 and haven't had to
replace a drive in well over 3 years. Your 30% failure rate is
ridiculously high.

MT

"Patrick" wrote in message
om...
Ablang wrote:
How long are hard drives supposed to last anyway (in your
experience)? I know some probably have an MTBF of about 30 years, but I
had a WD 6.x GB HD that just became as good as dead a few weeks ago (too
many bad sectors; scandisk ran forever).

So tell me about the bad ones in your life?

IDE drives are NOT intended to run over 50% service (on for 12 hours,
OFF for 12 hours!)

Really thats pretty much how my computer has been run over the last
5 years, on daytime, off at night, just like 99% of home computers.
Perhaps you had better inform the main PC manufacturers that
they have been using the wrong type of drives in their PC's for
the past decade or so.
It will be quite a shock to them!!


My best drives are SCSI, and the 10,000rpm Cheetah by Seagate last 7
years in 100% service, on servers! (my records). Zero failurs in 7 years.

Western Digital, Maxtor, Fujitsu, Hitachi, Seagate, and IBM IDE drives
seem to do OK on my 100% systems, but, I see from my records, that they
get to be RMA under warranty about 30%. So, about one out of three IDE
drives fails every two years!

The failed drives start doing a slow click, click click...
That lets me know that it is time to backup, R&R...
I run SCSI on these systems, also, so I jsut move the stuff over to a
SCSI drive!