PDA

View Full Version : MTBF question


DP
May 14th 06, 01:40 AM
Slightly off topic, but since there was a long discussion on this subject
under the heading "Dual Core Comparison," I thought I'd ask here first.

The MTBF number is good only if manufacturers are getting good reports from
the field about their products, right?

If my drive fails and it's out of warranty (or perhaps it's in warranty and
I can't find the paperwork to file the warranty claim) then I'm likely to
just get what data I can off the drive, destroy it and/or throw it away. As
a result, the manufacturer never knows the lifetime of my drive (or
thousands, if not millions, of others).

Is MTBF just a number derived from lab testing? Obviously, it doesn't
include those people who've had equipment failures and never notified the
equipment manufacturer of the failure. How reliable then is "mean" under
those circumstances?

NoNoBadDog!
May 14th 06, 09:21 AM
MTBF is a statistical projection based upon a dataset that will vary from
manufacturer to manufacturer.

As far as drive failure goes, it is not a matter of if, but a matter of
when.

There are options; Seagate, with their 5 year warranty, are currently the
best drives going. Why settle for less? Western Digital drives run hot and
have a relative high failure rate. I would not use WD for any reason.

Samsung and Hitachi are hit and miss...if you get lucky and get a "good"
drive, they will perform well and last a long time. OTOH, Samsung and
Hitachi are both famous for drives that die early deaths.

Finally, you get what you pay for. That white box 500GB no name drive on
sale for $59 after rebates should be a clue...do not buy this drive!

If your data is worth keeping, it is worth keeping on a quality drive.
Quality costs.
Simple equation.

Bobby

"DP" > wrote in message
.. .
> Slightly off topic, but since there was a long discussion on this subject
> under the heading "Dual Core Comparison," I thought I'd ask here first.
>
> The MTBF number is good only if manufacturers are getting good reports
> from the field about their products, right?
>
> If my drive fails and it's out of warranty (or perhaps it's in warranty
> and I can't find the paperwork to file the warranty claim) then I'm likely
> to just get what data I can off the drive, destroy it and/or throw it
> away. As a result, the manufacturer never knows the lifetime of my drive
> (or thousands, if not millions, of others).
>
> Is MTBF just a number derived from lab testing? Obviously, it doesn't
> include those people who've had equipment failures and never notified the
> equipment manufacturer of the failure. How reliable then is "mean" under
> those circumstances?
>
>

Cal Vanize
May 14th 06, 03:12 PM
NoNoBadDog! wrote:

> MTBF is a statistical projection based upon a dataset that will vary from
> manufacturer to manufacturer.
>
> As far as drive failure goes, it is not a matter of if, but a matter of
> when.
>
> There are options; Seagate, with their 5 year warranty, are currently the
> best drives going. Why settle for less? Western Digital drives run hot and
> have a relative high failure rate. I would not use WD for any reason.


Western Digital, with their high reliability and large MTBF, are
currently the best drives going. Why settle for less. Seagate drives
run hot and have a relatively high failure rate. I would not use
Seagate for any reason

Besides WDs are very fast,


>
> Samsung and Hitachi are hit and miss...if you get lucky and get a "good"
> drive, they will perform well and last a long time. OTOH, Samsung and
> Hitachi are both famous for drives that die early deaths.
>
> Finally, you get what you pay for. That white box 500GB no name drive on
> sale for $59 after rebates should be a clue...do not buy this drive!
>
> If your data is worth keeping, it is worth keeping on a quality drive.
> Quality costs.
> Simple equation.

If your data is worth keeping, it iw worth backing up on a quality drive.

(to each their own.)

General Schvantzkoph
May 14th 06, 03:51 PM
On Sat, 13 May 2006 19:40:48 -0500, DP wrote:

> Slightly off topic, but since there was a long discussion on this subject
> under the heading "Dual Core Comparison," I thought I'd ask here first.
>
> The MTBF number is good only if manufacturers are getting good reports
> from the field about their products, right?
>
> If my drive fails and it's out of warranty (or perhaps it's in warranty
> and I can't find the paperwork to file the warranty claim) then I'm likely
> to just get what data I can off the drive, destroy it and/or throw it
> away. As a result, the manufacturer never knows the lifetime of my drive
> (or thousands, if not millions, of others).
>
> Is MTBF just a number derived from lab testing? Obviously, it doesn't
> include those people who've had equipment failures and never notified the
> equipment manufacturer of the failure. How reliable then is "mean" under
> those circumstances?


MTBFs are based on lab testing not on field failures. They can also be
calculated based on the MTBFs of the components in the device.

John Weiss
May 14th 06, 06:03 PM
"NoNoBadDog!" > wrote...
> MTBF is a statistical projection based upon a dataset that will vary from
> manufacturer to manufacturer.

It also varies from drive line to drive line within a single mfgr. I don't know
if it is still true, but MTBFs/error rates at Seagate used to be a factor of 10
better for the Cheetah SCSI line than the Barracuda IDE lines.


> There are options; Seagate, with their 5 year warranty, are currently the
> best drives going. Why settle for less? Western Digital drives run hot and
> have a relative high failure rate. I would not use WD for any reason.

What are the reported failure rates for Cheetah, Barracuda, Raptor, and other
WD?

Ed Light
May 14th 06, 07:07 PM
Some people have reported using lots of Samsungs with no trouble.

The P80 series is the quietest of all the current hd's, and runs cool.

I think that providing good air flow is crucial for hd life.
--
Ed Light

Smiley :-/
MS Smiley :-\

Send spam to the FTC at

Thanks, robots.

Bring the Troops Home:
http://bringthemhomenow.org

Fight Spam:
http://bluesecurity.com

DP
May 15th 06, 02:16 AM
> "NoNoBadDog!" > wrote...
>> Western Digital drives run hot and have a relative high failure rate.

Western Digital says there is a firmware problem that's reporting the wrong
temps for the drive, and that they are actually running cooler than what
various programs are reporting. I own a WD (as well as a Maxtor) and my WD
does report higher temps.

When you say they run hot, do you know of any failures associated with high
temps? Just asking.
If there are a lot of temperature-related failures, then WD is b.s.'ing us
(me).

NoNoBadDog!
May 15th 06, 04:57 PM
"Cal Vanize" > wrote in message
...
>
>
> NoNoBadDog! wrote:
>
>> MTBF is a statistical projection based upon a dataset that will vary from
>> manufacturer to manufacturer.
>>
>> As far as drive failure goes, it is not a matter of if, but a matter of
>> when.
>>
>> There are options; Seagate, with their 5 year warranty, are currently
>> the best drives going. Why settle for less? Western Digital drives run
>> hot and have a relative high failure rate. I would not use WD for any
>> reason.
>
>
> Western Digital, with their high reliability and large MTBF, are currently
> the best drives going. Why settle for less. Seagate drives run hot and
> have a relatively high failure rate. I would not use Seagate for any
> reason
>
> Besides WDs are very fast,
>
>


This goes completely against every test published. Every one. In every
test I can find, WD is the hottest of *all* tested drives. 70% of the bad
drives I replace are WD.

WD are noisy, and are definitely not the fastest. Again, if you check the
tests online, you'll find many drives faster than WD.




>>
>> Samsung and Hitachi are hit and miss...if you get lucky and get a "good"
>> drive, they will perform well and last a long time. OTOH, Samsung and
>> Hitachi are both famous for drives that die early deaths.
>>
>> Finally, you get what you pay for. That white box 500GB no name drive on
>> sale for $59 after rebates should be a clue...do not buy this drive!
>>
>> If your data is worth keeping, it is worth keeping on a quality drive.
>> Quality costs.
>> Simple equation.
>
> If your data is worth keeping, it iw worth backing up on a quality drive.
>
> (to each their own.)
>
>

NoNoBadDog!
May 15th 06, 05:03 PM
"DP" > wrote in message
. ..
>
>> "NoNoBadDog!" > wrote...
>>> Western Digital drives run hot and have a relative high failure rate.
>
> Western Digital says there is a firmware problem that's reporting the
> wrong temps for the drive, and that they are actually running cooler than
> what various programs are reporting. I own a WD (as well as a Maxtor) and
> my WD does report higher temps.
>
> When you say they run hot, do you know of any failures associated with
> high temps? Just asking.
> If there are a lot of temperature-related failures, then WD is b.s.'ing us
> (me).
>
>
>
>
>


Even if the diode reports a wrong temp, I can attest based on my own
personal experience that WD run hotter. I have an Antec P160 with
temperature probes. My Seagates report 33 degrees celsius operating temps,
using both the probes and with a utility called Motherboard Monitor. My
sole WD drive reports a temp of 41 degrees celsius, and MBM reports a temp
of 42 degrees celsius.

As posted earlier, I service computers. About 70% of the drives I replace
due to failure are WD; and no it is not because 70% of the drives installed
are WD.

WD simply run hotter and have a shorter lifespan. I have observed this for
years.

Bobby