upgrade advice

upgrade advice

[Max Waterman]

Hi,

I have a 8x200GB RAID5 array - 4 SATA and 4 EIDE. It has 2 spare, so
that's only 6 drives actually in the array, making 1TB of space (I
think that's right).

The drives are quite old now, and people are asking what I want for
Christmas...so I'm wondering how to upgrade this setup. Clearly I
don't want to retire my existing drives prematurely, since they've
shown no sign of problem.

My initial thought was to make a RAID1 out of a new 1TB disk and my
existing RAID5 array, but perhaps that doesn't make much sense...

Perhaps I could get a 400GB drive (or more likely 500GB) and make a
RAID1 with a couple of the existing 200GBs?

...but I'm just guessing and the purpose of this email is to get advice.

What do you guys think?

Max.

Re: upgrade advice

[Justin Piszcz]

On Tue, 16 Dec 2008, Max Waterman wrote:

> Hi,
>
> I have a 8x200GB RAID5 array - 4 SATA and 4 EIDE. It has 2 spare, so
> that's only 6 drives actually in the array, making 1TB of space (I
> think that's right).
>
> The drives are quite old now, and people are asking what I want for
> Christmas...so I'm wondering how to upgrade this setup. Clearly I
> don't want to retire my existing drives prematurely, since they've
> shown no sign of problem.
>
> My initial thought was to make a RAID1 out of a new 1TB disk and my
> existing RAID5 array, but perhaps that doesn't make much sense...
>
> Perhaps I could get a 400GB drive (or more likely 500GB) and make a
> RAID1 with a couple of the existing 200GBs?
>
> ...but I'm just guessing and the purpose of this email is to get advice.
>
> What do you guys think?
>
> Max.

Tough call, depends on what you want to accomplish?  Availability or mass 
storage space?  You could continue running like you are now and use a 1TiB 
disk as your "scratch" area and rsync the important data off to your 
RAID-5 nightly for example.

Another option (probably not a good one) take two of your 200s and make a 
RAID1, leave your RAID-5 as-is, buy another 200GiB as a spare and then 
with the rest of the $ buy something else, there are no good alternatives, 
those are pretty old disks you got there, but I agree if its been running 
fine, no reason to replace it.

Justin.

Re: upgrade advice

[Redeeman]

On Tue, 2008-12-16 at 06:44 -0500, Justin Piszcz wrote:
> 
> On Tue, 16 Dec 2008, Max Waterman wrote:
> 
> > Hi,
> >
> > I have a 8x200GB RAID5 array - 4 SATA and 4 EIDE. It has 2 spare, so
> > that's only 6 drives actually in the array, making 1TB of space (I
> > think that's right).
> >
> > The drives are quite old now, and people are asking what I want for
> > Christmas...so I'm wondering how to upgrade this setup. Clearly I
> > don't want to retire my existing drives prematurely, since they've
> > shown no sign of problem.
> >
> > My initial thought was to make a RAID1 out of a new 1TB disk and my
> > existing RAID5 array, but perhaps that doesn't make much sense...
> >
> > Perhaps I could get a 400GB drive (or more likely 500GB) and make a
> > RAID1 with a couple of the existing 200GBs?
> >
> > ...but I'm just guessing and the purpose of this email is to get advice.
> >
> > What do you guys think?
> >
> > Max.
> 
> Tough call, depends on what you want to accomplish?  Availability or mass 
> storage space?  You could continue running like you are now and use a 1TiB 
> disk as your "scratch" area and rsync the important data off to your 
> RAID-5 nightly for example.
> 
> Another option (probably not a good one) take two of your 200s and make a 
> RAID1, leave your RAID-5 as-is, buy another 200GiB as a spare and then 
> with the rest of the $ buy something else, there are no good alternatives, 
> those are pretty old disks you got there, but I agree if its been running 
> fine, no reason to replace it.
Hehe, i myself am in a somewhat similar position.. i have a 6x300gb
raid5 array.. And the disks appears perfectly fine through smart etc,
but on the other hand, a SINGLE new disk today can present me with the
same level of usable capacity very cheap, and ill bet using considerably
less power. Thus i have chosen to very soon retire this setup, and make
a replacement 8x1tb or 6x1.5tb raid6 for increased capacity and
safety :)

i'd recommend you work towards deprecating LOTS of drives in favor of
fewer bigger newer, and keep the old ones around as an archive/backup
thing.

> 
> Justin.
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: upgrade advice

[Justin Piszcz]

On Tue, 16 Dec 2008, Redeeman wrote:

> On Tue, 2008-12-16 at 06:44 -0500, Justin Piszcz wrote:
>>
>> On Tue, 16 Dec 2008, Max Waterman wrote:
>>
>>> Hi,
>>>
>>> I have a 8x200GB RAID5 array - 4 SATA and 4 EIDE. It has 2 spare, so
>>> that's only 6 drives actually in the array, making 1TB of space (I
>>> think that's right).
>>>
>>> The drives are quite old now, and people are asking what I want for
>>> Christmas...so I'm wondering how to upgrade this setup. Clearly I
>>> don't want to retire my existing drives prematurely, since they've
>>> shown no sign of problem.
>>>
>>> My initial thought was to make a RAID1 out of a new 1TB disk and my
>>> existing RAID5 array, but perhaps that doesn't make much sense...
>>>
>>> Perhaps I could get a 400GB drive (or more likely 500GB) and make a
>>> RAID1 with a couple of the existing 200GBs?
>>>
>>> ...but I'm just guessing and the purpose of this email is to get advice.
>>>
>>> What do you guys think?
>>>
>>> Max.
>>
>> Tough call, depends on what you want to accomplish?  Availability or mass
>> storage space?  You could continue running like you are now and use a 1TiB
>> disk as your "scratch" area and rsync the important data off to your
>> RAID-5 nightly for example.
>>
>> Another option (probably not a good one) take two of your 200s and make a
>> RAID1, leave your RAID-5 as-is, buy another 200GiB as a spare and then
>> with the rest of the $ buy something else, there are no good alternatives,
>> those are pretty old disks you got there, but I agree if its been running
>> fine, no reason to replace it.
> Hehe, i myself am in a somewhat similar position.. i have a 6x300gb
> raid5 array.. And the disks appears perfectly fine through smart etc,
> but on the other hand, a SINGLE new disk today can present me with the
> same level of usable capacity very cheap, and ill bet using considerably
> less power. Thus i have chosen to very soon retire this setup, and make
> a replacement 8x1tb or 6x1.5tb raid6 for increased capacity and
> safety :)
>
> i'd recommend you work towards deprecating LOTS of drives in favor of
> fewer bigger newer, and keep the old ones around as an archive/backup
> thing.
Agree here, but the 1.5TiB are not enterprise disks, I would not recommend 
moving to them yet, but rather 1TiB, that is what I am currently doing, 
moving to WD RE3's (1TiB model).

Justin.

Re: upgrade advice

[Redeeman]

On Tue, 2008-12-16 at 08:07 -0500, Justin Piszcz wrote:
> 
> On Tue, 16 Dec 2008, Redeeman wrote:
> 
> > On Tue, 2008-12-16 at 06:44 -0500, Justin Piszcz wrote:
> >>
> >> On Tue, 16 Dec 2008, Max Waterman wrote:
> >>
> >>> Hi,
> >>>
> >>> I have a 8x200GB RAID5 array - 4 SATA and 4 EIDE. It has 2 spare, so
> >>> that's only 6 drives actually in the array, making 1TB of space (I
> >>> think that's right).
> >>>
> >>> The drives are quite old now, and people are asking what I want for
> >>> Christmas...so I'm wondering how to upgrade this setup. Clearly I
> >>> don't want to retire my existing drives prematurely, since they've
> >>> shown no sign of problem.
> >>>
> >>> My initial thought was to make a RAID1 out of a new 1TB disk and my
> >>> existing RAID5 array, but perhaps that doesn't make much sense...
> >>>
> >>> Perhaps I could get a 400GB drive (or more likely 500GB) and make a
> >>> RAID1 with a couple of the existing 200GBs?
> >>>
> >>> ...but I'm just guessing and the purpose of this email is to get advice.
> >>>
> >>> What do you guys think?
> >>>
> >>> Max.
> >>
> >> Tough call, depends on what you want to accomplish?  Availability or mass
> >> storage space?  You could continue running like you are now and use a 1TiB
> >> disk as your "scratch" area and rsync the important data off to your
> >> RAID-5 nightly for example.
> >>
> >> Another option (probably not a good one) take two of your 200s and make a
> >> RAID1, leave your RAID-5 as-is, buy another 200GiB as a spare and then
> >> with the rest of the $ buy something else, there are no good alternatives,
> >> those are pretty old disks you got there, but I agree if its been running
> >> fine, no reason to replace it.
> > Hehe, i myself am in a somewhat similar position.. i have a 6x300gb
> > raid5 array.. And the disks appears perfectly fine through smart etc,
> > but on the other hand, a SINGLE new disk today can present me with the
> > same level of usable capacity very cheap, and ill bet using considerably
> > less power. Thus i have chosen to very soon retire this setup, and make
> > a replacement 8x1tb or 6x1.5tb raid6 for increased capacity and
> > safety :)
> >
> > i'd recommend you work towards deprecating LOTS of drives in favor of
> > fewer bigger newer, and keep the old ones around as an archive/backup
> > thing.
> Agree here, but the 1.5TiB are not enterprise disks, I would not recommend 
> moving to them yet, but rather 1TiB, that is what I am currently doing, 
> moving to WD RE3's (1TiB model).
I am somewhat scared of the seagate stuff too.. I however will probably
go the WD10EADS way instead, i know its not the "enterprise" disks, but
they are really lowpower, and well.. i would hope raid6 can bail me out
of any disk failures.

> 
> Justin.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: upgrade advice

[Jon Nelson]

What about Hitachi?

Samsung has been my go-to drive for the better part of a decade, and
Seagate hasn't won any favor with me, so I've been relying on Samsung
and Hitachi with great results - so far.


-- 
Jon

Re: upgrade advice

[Ryan Wagoner]

Just buy two 1TB drives and retired your 200GB drives. Going from 8 to
2 drives will save you money in power costs and the speed of the 1TB
drives will be faster.

I second the Hitachi drives. I have 5 of the 1TB drives that are working great.

On Tue, Dec 16, 2008 at 10:57 AM, Jon Nelson
<jnelson-linux-raid@jamponi.net> wrote:
> What about Hitachi?
>
> Samsung has been my go-to drive for the better part of a decade, and
> Seagate hasn't won any favor with me, so I've been relying on Samsung
> and Hitachi with great results - so far.
>
>
> --
> Jon
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: upgrade advice

[Redeeman]

On Tue, 2008-12-16 at 09:57 -0600, Jon Nelson wrote:
> What about Hitachi?
> 
> Samsung has been my go-to drive for the better part of a decade, and
Well.. i had seriously considered the 1tb samsung, however, it appears
some people here havent had too good experience with it.. but i guess
there will always be some unlucky people. But the western digitals
appears to be the ones consuming lowest amount of power by far.

> Seagate hasn't won any favor with me, so I've been relying on Samsung
i have some 500 and 1000gb seagates in raid which works excellently.
> and Hitachi with great results - so far.
No experiences there, but looking at the power consumptions, i'd rather
not go there :)
> 
> 

Re: upgrade advice

[Justin Piszcz]

On Tue, 16 Dec 2008, Jon Nelson wrote:

> What about Hitachi?
>
> Samsung has been my go-to drive for the better part of a decade, and
> Seagate hasn't won any favor with me, so I've been relying on Samsung
> and Hitachi with great results - so far.
>
>
> -- 
> Jon
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Enterprise or 'desktop' disks Jon?

Justin.

Re: upgrade advice

[Jon Nelson]

On Tue, Dec 16, 2008 at 5:29 PM, Justin Piszcz <jpiszcz@lucidpixels.com> wrote:
>
>
> On Tue, 16 Dec 2008, Jon Nelson wrote:
>
>> What about Hitachi?
>>
>> Samsung has been my go-to drive for the better part of a decade, and
>> Seagate hasn't won any favor with me, so I've been relying on Samsung
>> and Hitachi with great results - so far.

> Enterprise or 'desktop' disks Jon?

Desktop. This is what I've got:

Hitachi Deskstar T7K500 (HDT725032VLA360)
SAMSUNG SpinPoint T166 series (HD321KJ)
Seagate Barracuda 7200.10 family (ST3320620AS)*
WDC WD3200AAKS-75VYA0

* the seagate has the infamous 3.AAK firmware.

I've almost always used Hitachi or Samsung, although I often
"repurpose" drives from friends and I've had quite a few Maxtor and
Seagate. The Hitachi does seem to run a bit hotter but it is probably
my favorite drive of the bunch. The Samsung is nice and quiet,
vibration-free, like the Hitachi and WD.  The Seagate vibrates and is
the second after the first died within the first few hours.

If I were to buy replacement drives, I almost always look at Hitachi
and Samsung first.

-- 
Jon

Re: upgrade advice

[Redeeman]

On Tue, 2008-12-16 at 18:08 -0600, Jon Nelson wrote:
> On Tue, Dec 16, 2008 at 5:29 PM, Justin Piszcz <jpiszcz@lucidpixels.com> wrote:
> >
> >
> > On Tue, 16 Dec 2008, Jon Nelson wrote:
> >
> >> What about Hitachi?
> >>
> >> Samsung has been my go-to drive for the better part of a decade, and
> >> Seagate hasn't won any favor with me, so I've been relying on Samsung
> >> and Hitachi with great results - so far.
> 
> > Enterprise or 'desktop' disks Jon?
> 
> Desktop. This is what I've got:
> 
> Hitachi Deskstar T7K500 (HDT725032VLA360)
> SAMSUNG SpinPoint T166 series (HD321KJ)
> Seagate Barracuda 7200.10 family (ST3320620AS)*
> WDC WD3200AAKS-75VYA0
> 
> * the seagate has the infamous 3.AAK firmware.
> 
> I've almost always used Hitachi or Samsung, although I often
> "repurpose" drives from friends and I've had quite a few Maxtor and
> Seagate. The Hitachi does seem to run a bit hotter but it is probably
> my favorite drive of the bunch. The Samsung is nice and quiet,
> vibration-free, like the Hitachi and WD.  The Seagate vibrates and is
> the second after the first died within the first few hours.
> 
> If I were to buy replacement drives, I almost always look at Hitachi
> and Samsung first.

What sort of volume of disks are you using, and what loads? (24/7 with
high load?)

> 

Re: upgrade advice

[Jon Nelson]

> What sort of volume of disks are you using, and what loads? (24/7 with
> high load?)

It's a home server. It's up 24/7. Load probably 80% of the time is
low, the rest of the time it's bursty.

-- 
Jon

RE: upgrade advice

[David Lethe]

> -----Original Message-----
> From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-
> owner@vger.kernel.org] On Behalf Of Jon Nelson
> Sent: Wednesday, December 17, 2008 7:27 AM
> To: Redeeman
> Cc: Justin Piszcz; linux-raid@vger.kernel.org
> Subject: Re: upgrade advice
> 
> > What sort of volume of disks are you using, and what loads? (24/7
> with
> > high load?)
> 
> It's a home server. It's up 24/7. Load probably 80% of the time is
> low, the rest of the time it's bursty.
> 
> --
> Jon
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid"
> in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Read the specs on the disks.  Most consumer class drives are rated for
only 2400 hours annual duty cycle ... so I guess you turn the computer
off in April? :)

Other differences include number of ECC correction bits, so you will
absolutely get more grown bad blocks with cheap drives.

Consider also how many people complain about botched rebuilds due to
multiple drive failures on rebuilds and bad blocks on surviving disks.
I can't remember anybody who bought enterprise class disks asking for
such help recently, it always seems to be people who buy consumer
drives.   No wonder several fail within days of each other, they all
have same model, I/O load, and generally same manufacturing batch.

If you are hell-bent on getting cheap drives, then at least factor in
cost of an additional drive so you can implement RAID6, and automate a
daemon to check/repair consistency often.

Life is short.  Spend the extra money and get disks designed to run in
servers, not PCs. Unless, of course, you have a rather large autochanger
and love to use it.


David
 

Re: upgrade advice

[Jon Nelson]

On Wed, Dec 17, 2008 at 8:28 AM, David Lethe <david@santools.com> wrote:
>> -----Original Message-----
>> From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-
>> owner@vger.kernel.org] On Behalf Of Jon Nelson
>> Sent: Wednesday, December 17, 2008 7:27 AM
>> To: Redeeman
>> Cc: Justin Piszcz; linux-raid@vger.kernel.org
>> Subject: Re: upgrade advice
>>
>> > What sort of volume of disks are you using, and what loads? (24/7
>> with
>> > high load?)
>>
>> It's a home server. It's up 24/7. Load probably 80% of the time is
>> low, the rest of the time it's bursty.
>>

> Read the specs on the disks.  Most consumer class drives are rated for
> only 2400 hours annual duty cycle ... so I guess you turn the computer
> off in April? :)

Yeah. Right. I typically get > 5 years out of the disks. I got almost
9 years once, before bad sectors started showing up. Actually, I've
either gotten less than a week or more than 4 years out of every
single drive I've ever had, except a really bad batch of seagates I
got 8-10 years ago.

The current temp of the drives varies between 28C and 33C (the Hitach
is warmer by +4C than any other drive).

> Other differences include number of ECC correction bits, so you will
> absolutely get more grown bad blocks with cheap drives.

That's good to know.

> drives.   No wonder several fail within days of each other, they all
> have same model, I/O load, and generally same manufacturing batch.

I never have more than 1 of the same manuf. / model in a raid at a
time. I have a 3 drive raid10f2 with 3 different manuf.

> If you are hell-bent on getting cheap drives, then at least factor in
> cost of an additional drive so you can implement RAID6, and automate a
> daemon to check/repair consistency often.

I will likely move to raid6 eventually.
Thanks for the advice.

-- 
Jon

RE: upgrade advice

[David Lethe]

> -----Original Message-----
> From: Jon Nelson [mailto:jnelson-linux-raid@jamponi.net]
> Sent: Wednesday, December 17, 2008 8:38 AM
> To: David Lethe
> Cc: Redeeman; Justin Piszcz; linux-raid@vger.kernel.org
> Subject: Re: upgrade advice
> 
> On Wed, Dec 17, 2008 at 8:28 AM, David Lethe <david@santools.com>
> wrote:
> >> -----Original Message-----
> >> From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-
> >> owner@vger.kernel.org] On Behalf Of Jon Nelson
> >> Sent: Wednesday, December 17, 2008 7:27 AM
> >> To: Redeeman
> >> Cc: Justin Piszcz; linux-raid@vger.kernel.org
> >> Subject: Re: upgrade advice
> >>
> >> > What sort of volume of disks are you using, and what loads? (24/7
> >> with
> >> > high load?)
> >>
> >> It's a home server. It's up 24/7. Load probably 80% of the time is
> >> low, the rest of the time it's bursty.
> >>
> 
> > Read the specs on the disks.  Most consumer class drives are rated
> for
> > only 2400 hours annual duty cycle ... so I guess you turn the
> computer
> > off in April? :)
> 
> Yeah. Right. I typically get > 5 years out of the disks. I got almost
> 9 years once, before bad sectors started showing up. Actually, I've
> either gotten less than a week or more than 4 years out of every
> single drive I've ever had, except a really bad batch of seagates I
> got 8-10 years ago.
> 
> The current temp of the drives varies between 28C and 33C (the Hitach
> is warmer by +4C than any other drive).
> 
> > Other differences include number of ECC correction bits, so you will
> > absolutely get more grown bad blocks with cheap drives.
> 
> That's good to know.
> 
> > drives.   No wonder several fail within days of each other, they all
> > have same model, I/O load, and generally same manufacturing batch.
> 
> I never have more than 1 of the same manuf. / model in a raid at a
> time. I have a 3 drive raid10f2 with 3 different manuf.
> 
> > If you are hell-bent on getting cheap drives, then at least factor
in
> > cost of an additional drive so you can implement RAID6, and automate
> a
> > daemon to check/repair consistency often.
> 
> I will likely move to raid6 eventually.
> Thanks for the advice.
> 
> --
> Jon
The duty cycle makes a difference now, but wasn't a design point until
last few years ago when hardware RAID manufacturers reluctantly started
supporting SATA disks. When you were buying disks 5-10 years ago, the
enterprise class disks were SCSI & FC, and consumer drives were
ATA/SATA. They had to charge more and make the ATA disks a little better
to prevent losing too much money to warranty replacements.  The market
now demands clear pricing and performance/reliability differences
between enterprise and consumer class SATA devices.

As for drive temperature ... believe it or not, it is irrelevant when it
comes to drive failures, unless you are pushing the operational
temperature thresholds.  Google published a detailed analysis of drive
failures in their storage farm that included average drive temperatures,
and proved that increased drive temp did not affect failure rate.   (In
fact, the drives running at the lowest temperatures actually had a
slightly higher failure rate).

Anybody who thinks there are no difference between the specs haven't
looked at them deeply enough.  ECC bits;  background media
scanning/repair algorithms; and firmware make all the difference in the
world.  Ask anybody who has worked as a storage architect for a RAID
manufacturer or is a drive test engineer. 

To the untrained eye (or somebody who has never had opportunity to
attend non-disclosure meetings with drive manufacturers), there isn't
much of a difference because they get hung up on easy things like RPMs
and amount of cache .. the stuff they put on the outside of the box.

Error detection, recovery algorithms, extensive log page reporting, and
online/offline firmware diagnostics tend to be ignored. Not only do
people not understand them, but some of the really good stuff isn't
published in the manuals due to intellectual property concerns.  If you
are in the 'biz, and buying thousands or tens of thousands of disks a
month, you become well aware of these things.

"Well, I buy xyz brand disk drives because I had such-and-such
experiences with abc brand disks".   How many times have people said
that??  The same person wouldn't commoditize car, dishwashers, or wine
like this.  People who don't understand a product say such things.   

I should touch on firmware as well.  NCQ, TCQ, severety-1 bugs that can
result in disks locking up or having catastrophic data loss?   They
exist.  I won't break any NDAs, but even a firmware upgrade can have
profound differences in your storage farm.   The firmware update release
notes for many drives would scare the heck out of you, and make you
wonder what motivated ABC company to ever release something, or for that
matter, what motivated ABC company to not make something public and have
a massive recall when certain bugs were found.   We've seen this happen
in industry before with things like Seagate's stiction problem, NCQ
bugs, IBM "deathstars", Hitachi recalls and so on.

OK, off my soapbox ... back to work writing disk diagnostic software for
my OEM customers 

David

RE: upgrade advice

[Redeeman]

On Wed, 2008-12-17 at 09:30 -0600, David Lethe wrote:
> > -----Original Message-----
> > From: Jon Nelson [mailto:jnelson-linux-raid@jamponi.net]
> > Sent: Wednesday, December 17, 2008 8:38 AM
> > To: David Lethe
> > Cc: Redeeman; Justin Piszcz; linux-raid@vger.kernel.org
> > Subject: Re: upgrade advice
> > 
> > On Wed, Dec 17, 2008 at 8:28 AM, David Lethe <david@santools.com>
> > wrote:
> > >> -----Original Message-----
> > >> From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-
> > >> owner@vger.kernel.org] On Behalf Of Jon Nelson
> > >> Sent: Wednesday, December 17, 2008 7:27 AM
> > >> To: Redeeman
> > >> Cc: Justin Piszcz; linux-raid@vger.kernel.org
> > >> Subject: Re: upgrade advice
> > >>
> > >> > What sort of volume of disks are you using, and what loads? (24/7
> > >> with
> > >> > high load?)
> > >>
> > >> It's a home server. It's up 24/7. Load probably 80% of the time is
> > >> low, the rest of the time it's bursty.
> > >>
> > 
> > > Read the specs on the disks.  Most consumer class drives are rated
> > for
> > > only 2400 hours annual duty cycle ... so I guess you turn the
> > computer
> > > off in April? :)
> > 
> > Yeah. Right. I typically get > 5 years out of the disks. I got almost
> > 9 years once, before bad sectors started showing up. Actually, I've
> > either gotten less than a week or more than 4 years out of every
> > single drive I've ever had, except a really bad batch of seagates I
> > got 8-10 years ago.
> > 
> > The current temp of the drives varies between 28C and 33C (the Hitach
> > is warmer by +4C than any other drive).
> > 
> > > Other differences include number of ECC correction bits, so you will
> > > absolutely get more grown bad blocks with cheap drives.
> > 
> > That's good to know.
> > 
> > > drives.   No wonder several fail within days of each other, they all
> > > have same model, I/O load, and generally same manufacturing batch.
> > 
> > I never have more than 1 of the same manuf. / model in a raid at a
> > time. I have a 3 drive raid10f2 with 3 different manuf.
> > 
> > > If you are hell-bent on getting cheap drives, then at least factor
> in
> > > cost of an additional drive so you can implement RAID6, and automate
> > a
> > > daemon to check/repair consistency often.
> > 
> > I will likely move to raid6 eventually.
> > Thanks for the advice.
> > 
> > --
> > Jon
> The duty cycle makes a difference now, but wasn't a design point until
> last few years ago when hardware RAID manufacturers reluctantly started
> supporting SATA disks. When you were buying disks 5-10 years ago, the
> enterprise class disks were SCSI & FC, and consumer drives were
> ATA/SATA. They had to charge more and make the ATA disks a little better
> to prevent losing too much money to warranty replacements.  The market
> now demands clear pricing and performance/reliability differences
> between enterprise and consumer class SATA devices.
> 
> As for drive temperature ... believe it or not, it is irrelevant when it
> comes to drive failures, unless you are pushing the operational
> temperature thresholds.  Google published a detailed analysis of drive
> failures in their storage farm that included average drive temperatures,
> and proved that increased drive temp did not affect failure rate.   (In
> fact, the drives running at the lowest temperatures actually had a
> slightly higher failure rate).
> 
> Anybody who thinks there are no difference between the specs haven't
> looked at them deeply enough.  ECC bits;  background media
> scanning/repair algorithms; and firmware make all the difference in the
> world.  Ask anybody who has worked as a storage architect for a RAID
> manufacturer or is a drive test engineer. 
> 
> To the untrained eye (or somebody who has never had opportunity to
> attend non-disclosure meetings with drive manufacturers), there isn't
> much of a difference because they get hung up on easy things like RPMs
> and amount of cache .. the stuff they put on the outside of the box.
> 
> Error detection, recovery algorithms, extensive log page reporting, and
> online/offline firmware diagnostics tend to be ignored. Not only do
> people not understand them, but some of the really good stuff isn't
> published in the manuals due to intellectual property concerns.  If you
> are in the 'biz, and buying thousands or tens of thousands of disks a
> month, you become well aware of these things.
> 
> "Well, I buy xyz brand disk drives because I had such-and-such
> experiences with abc brand disks".   How many times have people said
> that??  The same person wouldn't commoditize car, dishwashers, or wine
> like this.  People who don't understand a product say such things.   
Yeah i realize that this is really a useless thing. Im sure these people
have had the failures they speak of, but thats ofcourse totally useless
across series of disks, and ALL brands and series obviously will have
some disks that fail early, and someone is bound to get those..

> 
> I should touch on firmware as well.  NCQ, TCQ, severety-1 bugs that can
> result in disks locking up or having catastrophic data loss?   They
> exist.  I won't break any NDAs, but even a firmware upgrade can have
> profound differences in your storage farm.   The firmware update release
> notes for many drives would scare the heck out of you, and make you
> wonder what motivated ABC company to ever release something, or for that
> matter, what motivated ABC company to not make something public and have
> a massive recall when certain bugs were found.   We've seen this happen
> in industry before with things like Seagate's stiction problem, NCQ
> bugs, IBM "deathstars", Hitachi recalls and so on.
> 


Hmm.. you have kindof convinced me.. I should probably look at buying WD
RE2-GP then


> OK, off my soapbox ... back to work writing disk diagnostic software for
> my OEM customers 
> 
> David
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: upgrade advice

[John Robinson]

On 17/12/2008 15:30, David Lethe wrote:
> The duty cycle makes a difference now, but wasn't a design point until
[...]
> OK, off my soapbox ... back to work writing disk diagnostic software for
> my OEM customers 

I am going to print that message, frame it and hang it on the wall. 
(Including the bit I elided with ...)

I'll probably keep using desktop drives for domestic NAS, though.

Cheers,

John.

Re: upgrade advice

[Justin Piszcz]

On Thu, 18 Dec 2008, John Robinson wrote:

> On 17/12/2008 15:30, David Lethe wrote:
>> The duty cycle makes a difference now, but wasn't a design point until
> [...]
>> OK, off my soapbox ... back to work writing disk diagnostic software for
>> my OEM customers 
>
> I am going to print that message, frame it and hang it on the wall. 
> (Including the bit I elided with ...)
>
> I'll probably keep using desktop drives for domestic NAS, though.
I think it also depends on how often drives are used and what type of 
workloads they are exposed to, for a domestic NAS for mainly sequential 
file writes/reads, they will probably be OK-- I have a few SW RAID5's on 
desktop drives but I use them primarily as storage via rsync, once a week 
or month I am not constantly read/writing on them-- on a daily basis.  Is
that how you use your NAS as well?  Or?

Re: upgrade advice

[John Robinson]

On 18/12/2008 16:43, Justin Piszcz wrote:
> On Thu, 18 Dec 2008, John Robinson wrote:
>> On 17/12/2008 15:30, David Lethe wrote:
>>> The duty cycle makes a difference now, but wasn't a design point until
>> [...]
>>> OK, off my soapbox ... back to work writing disk diagnostic software for
>>> my OEM customers 
>>
>> I am going to print that message, frame it and hang it on the wall. 
>> (Including the bit I elided with ...)
>>
>> I'll probably keep using desktop drives for domestic NAS, though.
> I think it also depends on how often drives are used and what type of 
> workloads they are exposed to, for a domestic NAS for mainly sequential 
> file writes/reads, they will probably be OK-- I have a few SW RAID5's on 
> desktop drives but I use them primarily as storage via rsync, once a 
> week or month I am not constantly read/writing on them-- on a daily 
> basis.  Is that how you use your NAS as well?  Or?

I have 2 scenarios. For myself, I've a big box I put together myself, 
with a P45-chipset mobo, a Core 2 Quad, 4GB RAM (so far), a Supermicro 
hot-swap SATA enclosure and 3 1TB Samsungs in Linux RAID-5, which serves 
both my IT support work where I work from home (storing ISO images, 
running various Windows and Linux VMs mostly for testing/development) 
and my domestic NAS demands (MythTV server, SqueezeCenter, and backups 
of other PCs). It's on 24/7 but there's only me to make demands of it so 
it doesn't get very heavily exercised. I'll probably replace the current 
Samsungs (which have consecutive serial numbers) over time, even if they 
aren't failing, with others from other batches and other makes. I always 
recommend my business clients use enterprise drives in any central 
storage application (but it seems I don't take my own advice when it 
comes to myself).

I am also involved in an assisting capacity with a group of home and 
multiroom hifi folks, who need NASes for storing the music to stream to 
TwonkyMedia, SqueezeCenter etc. We like the ReadyNAS range (Linux RAID 
based), and have tried Seagate desktop and enterprise drives as well as 
the Samsung desktop drives. So far so good, no problems, and we like the 
Samsungs best so far, because they're quieter and have been cheaper. The 
ReadyNAS knows how to sleep, so desktop drives should be fine here as 
they'll be shut down when people aren't playing their hifi (though 
there's a bug in current versions of SqueezeCenter that stops the drives 
sleeping if your Duet remote's turned on) and they're rarely doing 
anything strenuous.

As always, I'd appreciate comments from people better informed than me!

Cheers,

John.

PS. Who mentioned big autochangers? Somebody did. Anyway, I just bought 
a Dell PowerVault 132T off eBay with an LTO-2 drive for next to nothing 
:-) so I shall have many happy hours backing up my 2TB array onto lots 
of tapes.

RE: upgrade advice / Disk drive failure rates - real world

[David Lethe]

> -----Original Message-----
> From: Justin Piszcz [mailto:jpiszcz@lucidpixels.com]
> Sent: Thursday, December 18, 2008 10:43 AM
> To: John Robinson
> Cc: David Lethe; Linux RAID
> Subject: Re: upgrade advice
> 
> 
> 
> On Thu, 18 Dec 2008, John Robinson wrote:
> 
> > On 17/12/2008 15:30, David Lethe wrote:
> >> The duty cycle makes a difference now, but wasn't a design point
> until
> > [...]
> >> OK, off my soapbox ... back to work writing disk diagnostic
software
> for
> >> my OEM customers
> >
> > I am going to print that message, frame it and hang it on the wall.
> > (Including the bit I elided with ...)
> >
> > I'll probably keep using desktop drives for domestic NAS, though.
> I think it also depends on how often drives are used and what type of
> workloads they are exposed to, for a domestic NAS for mainly
sequential
> file writes/reads, they will probably be OK-- I have a few SW RAID5's
> on
> desktop drives but I use them primarily as storage via rsync, once a
> week
> or month I am not constantly read/writing on them-- on a daily basis.
> Is
> that how you use your NAS as well?  Or?
> 

Surprisingly, the failure rates for consumer class ATA disks are
counterintuitive.
Breaking down into raw low,medium,high utilization ... this is what the
study of 100,000 disks
Revealed over 5 year period. (Rounded results, all in percentage)

DISK AGE      LOW   MED  HIGH (utilization)
 3 months      4     2    10
 6 mos         2     1     4
12 mos         .5    1     2
2  years       2     2     2
4  years       3    4      4
5  years       1    1      5

So, if you pound a disk with I/O then it is 5x more likely to die in
first 3 months then
if it has light duty load.   If a disk survives the first year, then
load doesn't make much of
a statistical difference ... until it approaches the 5 year mark.

Note, these are for consumer class disks, that were better products back
in 2000 when the study began.
There are no long-term studies for real-world drive life for the (SATA)
consumer vs. enterprise disk drives that are made today.   The study
involved 9 different disks, Seagate, Hitachi, WD, etc.   that were
typical of what you got with a personal computer from manufacturer, or
you bought at a PC store.   The test is as real-world as they come.

As for drive temperature vs. failure rate, this will blow you away.

The probability density curve is logarithmic in nature, so disks at 20
degrees C are 3X more likely to fail for any given load then disks at 26
degrees C.   Once you hit 26 degrees then the curve really flattens out.
Not much of a difference between 30-45 degrees.   Sweet spot is 36-42
degrees.  Colder disks are 6x more likely to fail then ones running
around 40 deg.

So don't throw money away on disk drive coolers/fans, unless temp
without them is in the 45 deg C range!

There is more correlation between failure rates of drives when both age
and temperature are considered.   The highlight that disks running
colder than the sweet spot are 2-3x more likely to die in first 3 months
then ones running near 40 deg.  (AFR is 9% for coldest disks, 3% for
warmest disks).


From 6 mos to 2 years, drive temp becomes less of a factor, with AFR of
cold disks around 4%, sweet-spot disks 2%.
Year 3, however is the killer, where everything changes...
AFR of cold-warm disks approx 6 %, AFR of the disks from 35-40 deg is
11%.

So aggregate walk-away
 - Keep disks in 36-42 degrees C for maximum life up to year 2, where
they should be run cooler.
 - New disks are 5x more likely to die in first 3 months with high vs.
low workload.
 - AFR and temperature become less important after the 3-month burn-in,
but this all changes in year 3.

David @ santools.com

RE: upgrade advice / Disk drive failure rates - real world

[Max Waterman]

On Thu, 18 Dec 2008 13:14:27 -0600, "David Lethe" <david@santools.com>
said:
snip
>
> So aggregate walk-away
>  - Keep disks in 36-42 degrees C for maximum life up to year 2, where
> they should be run cooler.
>  - New disks are 5x more likely to die in first 3 months with high vs.
> low workload.

I wonder if you can use this fact to make drive which are more likely to
fail later, actually fail sooner.

I'm thinking that I could run my new drives in a fridge for a month or
two or three - if they fail, I'll return them, else I can be more sure
they'll last their expected life and not give me any trouble.

Does that make any sense?

Also, I see people recommending enterprise disks, which makes me think
that those same people have forgotten what the 'I' stands for in 'RAID'.
Of course, 'inexpensive' is relative - probably to one's income and it
seems some people are paid more than others....it almost suggests we
should have another acronym - RADCD - redundant array of dirt cheap
disks. :)

Max.

RE: upgrade advice / Disk drive failure rates - real world

[David Lethe]

> -----Original Message-----
> From: Max Waterman [mailto:davidmaxwaterman@fastmail.co.uk]
> Sent: Thursday, December 18, 2008 4:52 PM
> To: David Lethe; Justin Piszcz; John Robinson
> Cc: Linux RAID
> Subject: RE: upgrade advice / Disk drive failure rates - real world
> 
> 
> On Thu, 18 Dec 2008 13:14:27 -0600, "David Lethe" <david@santools.com>
> said:
> snip
> >
> > So aggregate walk-away
> >  - Keep disks in 36-42 degrees C for maximum life up to year 2, where
> > they should be run cooler.
> >  - New disks are 5x more likely to die in first 3 months with high
> vs.
> > low workload.
> 
> I wonder if you can use this fact to make drive which are more likely
> to
> fail later, actually fail sooner.
> 
> I'm thinking that I could run my new drives in a fridge for a month or
> two or three - if they fail, I'll return them, else I can be more sure
> they'll last their expected life and not give me any trouble.
> 
> Does that make any sense?
> 
> Also, I see people recommending enterprise disks, which makes me think
> that those same people have forgotten what the 'I' stands for in
> 'RAID'.
> Of course, 'inexpensive' is relative - probably to one's income and it
> seems some people are paid more than others....it almost suggests we
> should have another acronym - RADCD - redundant array of dirt cheap
> disks. :)
> 
> Max.

If it is your desire to fail the drive as quickly as possible, you definitely need that environmental chamber, but more importantly the right software test suite. The big-box RAID vendors spend a lot of money developing 
stress environments so they weed out the ones that are likely to fail before they ship, and the
code they use is a big trade secret because this differentiates the tier-1s (EMC, NetApp, etc..) from 
everybody else who doesn't have enclosures that can survive a shotgun blast or a drop test.  Th

It is public knowledge that sudden changes in temperature under high IOPs & seeks is a reasonable part of 
a test suite.  Large block copies isn't stressful.  You will want to change some of the programmable drive
settings, but won't tell you specifics because I am not at liberty to tell you what any particular vendor does.
(But some settings make a lot of sense if you read the ANSI specs and drive vendor's product manuals and you put 2+2
Together).

Vibration is good too.  Don't use your freezer however, there are better ways to go.   For instance, if you were lucky enough to get an Omaha steaks gift package like me yesterday, then you have a 100% Styrofoam chest that is perfect for your needs.  You can stick in some dry ice, carve out a hole for a hair dryer to fit into, and put in a fan in the bottom so you can equalize pressure.    You will want to create some vibrations.  Maybe use a massage pad, or you can easily create enough sonic pressure if you have a teenage daughter with a cell phone.

It won't be nearly as good as a tier-1.  They just need 48 hours with their equipment to do better than what you can get with a 3 months worth of usage.    Still perhaps a week in the Styrofoam ice chest with random small block I/O
Will get you close enough.   

David

RE: upgrade advice

[Redeeman]

On Wed, 2008-12-17 at 08:28 -0600, David Lethe wrote:
> > -----Original Message-----
> > From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-
> > owner@vger.kernel.org] On Behalf Of Jon Nelson
> > Sent: Wednesday, December 17, 2008 7:27 AM
> > To: Redeeman
> > Cc: Justin Piszcz; linux-raid@vger.kernel.org
> > Subject: Re: upgrade advice
> > 
> > > What sort of volume of disks are you using, and what loads? (24/7
> > with
> > > high load?)
> > 
> > It's a home server. It's up 24/7. Load probably 80% of the time is
> > low, the rest of the time it's bursty.
> > 
> > --
> > Jon
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-raid"
> > in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> Read the specs on the disks.  Most consumer class drives are rated for
> only 2400 hours annual duty cycle ... so I guess you turn the computer
> off in April? :)
> 
> Other differences include number of ECC correction bits, so you will
> absolutely get more grown bad blocks with cheap drives.
> 
> Consider also how many people complain about botched rebuilds due to
> multiple drive failures on rebuilds and bad blocks on surviving disks.
> I can't remember anybody who bought enterprise class disks asking for
> such help recently, it always seems to be people who buy consumer
> drives.   No wonder several fail within days of each other, they all
> have same model, I/O load, and generally same manufacturing batch.

Did you ever split open the consumer and "enterprise" versions of the
drives and observe?

While i do believe they do some more testing/quality control on the
enterprise disks, they are almost all identical excepting firmware.. so
to be fair, i'd say the mistake is not properly stress testing the
disks, and for instance just buying a huge batch of disks and putting to
use..

> 
> If you are hell-bent on getting cheap drives, then at least factor in
> cost of an additional drive so you can implement RAID6, and automate a
> daemon to check/repair consistency often.
> 
> Life is short.  Spend the extra money and get disks designed to run in
> servers, not PCs. Unless, of course, you have a rather large autochanger
> and love to use it.
> 
> 
> David
>  
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: upgrade advice

[Martin K. Petersen]

>>>>> "Redeeman" == Redeeman  <redeeman@metanurb.dk> writes:

>> Consider also how many people complain about botched rebuilds due
>> to multiple drive failures on rebuilds and bad blocks on surviving
>> disks.  I can't remember anybody who bought enterprise class disks
>> asking for such help recently, it always seems to be people who buy
>> consumer drives.  No wonder several fail within days of each other,
>> they all have same model, I/O load, and generally same
>> manufacturing batch.

Redeeman> Did you ever split open the consumer and "enterprise"
Redeeman> versions of the drives and observe?

Redeeman> While i do believe they do some more testing/quality control
Redeeman> on the enterprise disks, they are almost all identical
Redeeman> excepting firmware.. so to be fair, i'd say the mistake is
Redeeman> not properly stress testing the disks, and for instance just
Redeeman> buying a huge batch of disks and putting to use..

People always seem to assume that hardware is what's making the
difference between consumer and enterprise.  It's not.  The physical
hardware differs mostly due to capacity vs. RPM trade-offs.  Most
vendors these days have big platters for high-capacity drives and
smaller platters for high RPM/higher IOPS class drives.  On top of
that, head/platter count may vary in capacity classes within a series.

But the important difference between consumer and enterprise drives is
not mechanical.  It's the firmware.  Consumer drive firmware is about
squeezing out the most capacity/$ and nothing else.

Enterprise drives trade capacity for reliability by way of the
firmware.  That includes many things like using more space for track
info (gap/sync), much better ECC, better tolerance for rotational
vibration, etc.

Most of the errors you see on drives are a result of media errors that
are big enough that the drive ECC can't correct them.  Errors are
often caused by head misses due to bad tracking, vibration from other
drives in the enclosure, the user kicking the cabinet at an
inopportune moment, etc.  I.e. external interference.  Other errors
are due to real imperfections of the media itself.

Enterprise drive firmware is about being more resistant to outside
factors as well as real media defects.  That firmware cost more to
develop than the consumer ditto.  And the vendors charge a premium for
it.

So it's not the size that matters.  It's how you use it.

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: upgrade advice

[Bill Davidsen]

Martin K. Petersen wrote:
> People always seem to assume that hardware is what's making the
> difference between consumer and enterprise.  It's not.  The physical
> hardware differs mostly due to capacity vs. RPM trade-offs.  Most
> vendors these days have big platters for high-capacity drives and
> smaller platters for high RPM/higher IOPS class drives.  On top of
> that, head/platter count may vary in capacity classes within a series.
>
> But the important difference between consumer and enterprise drives is
> not mechanical.  It's the firmware.  Consumer drive firmware is about
> squeezing out the most capacity/$ and nothing else.
>
>   
That's simply not the case. Cost is one of the issues, but the typical 
use of the drive is one of the most important things about the firmware. 
With consumer drives, it's likely that this is the one and only drive 
holding the data, so clever retries in case of error are important. With 
server grade drives, it's likely they are in a RAID, so returning the 
error quickly so the controller or OS can compensate is the important 
issue. That's been discussed here before, some drives even give you a 
choice of "don't hang up" vs. "try like hell" on errors, through jumpers 
or firmware.

> Enterprise drives trade capacity for reliability by way of the
> firmware.  That includes many things like using more space for track
> info (gap/sync), much better ECC, better tolerance for rotational
> vibration, etc.
>
> Most of the errors you see on drives are a result of media errors that
> are big enough that the drive ECC can't correct them.  Errors are
> often caused by head misses due to bad tracking, vibration from other
> drives in the enclosure, the user kicking the cabinet at an
> inopportune moment, etc.  I.e. external interference.  Other errors
> are due to real imperfections of the media itself.
>
>   
I would be surprised if a consumer grade drive doing more retries over 
several seconds rather than several rotations wasn't better able to 
correct for most of the transient problems you mention. So your comments 
about transient mechanical issues aren't telling me much, other than 
server drives being more likely to get vibration from other drives.

> Enterprise drive firmware is about being more resistant to outside
> factors as well as real media defects.  That firmware cost more to
> develop than the consumer ditto.  And the vendors charge a premium for
> it.
>
>   
Other than possibly having more ECC bits there isn't much difference, as 
several people here have noted you don't want the drive to hang for 
several seconds trying this and that in a server environment. And given 
that there are a very small number of things to be done on error, like 
reread, seek away and back, recalibrate, etc, I would be amazed if 
vendors didn't just put all the code in the firmware and use a little 
table to determine which actions to take in what order, and how many 
times. The idea of some vast and complex code just doesn't fly, there 
aren't that many things to try.

-- 
Bill Davidsen <davidsen@tmr.com>
  "Woe unto the statesman who makes war without a reason that will still
  be valid when the war is over..." Otto von Bismark 

Re: upgrade advice

[Martin K. Petersen]

>>>>> "Bill" == Bill Davidsen <davidsen@tmr.com> writes:

>> Most of the errors you see on drives are a result of media errors
>> that are big enough that the drive ECC can't correct them.  Errors
>> are often caused by head misses due to bad tracking, vibration from
>> other drives in the enclosure, the user kicking the cabinet at an
>> inopportune moment, etc.  I.e. external interference.  Other errors
>> are due to real imperfections of the media itself.

Bill> I would be surprised if a consumer grade drive doing more retries
Bill> over several seconds rather than several rotations wasn't better
Bill> able to correct for most of the transient problems you mention.

Not all the problems I mentioned are of transient nature.  Several
common corruption scenarios are caused by the transient external factors
*at write time*.  No amount of retrying is going to fix something that
was badly written to begin with.  Doesn't even have to be the sector in
question.  Could be adjacent tracks that got clobbered.


Bill> Other than possibly having more ECC bits there isn't much
Bill> difference,

I mentioned better tracking/multiple sync marks as another crucial
difference.  That's a pretty huge deal in my book.

Nearline drive firmware also devotes resources to predicting impending
failure.  They have the ability to throttle the I/O pipeline if there's
an increased risk of write error due to excessive seeking, overheating,
etc.  That means that under load performance can be choppy.

That is unacceptable behavior in the consumer/interactivity
benchmarketing-focused market whereas making sure you write things
correctly is an absolute must in the enterprise space.  And the
non-deterministic performance characteristics are not such a big deal
when the drives are sitting behind an array head with non-volatile
cache.


Bill> as several people here have noted you don't want the drive to hang
Bill> for several seconds trying this and that in a server
Bill> environment. And given that there are a very small number of
Bill> things to be done on error, like reread, seek away and back,
Bill> recalibrate, etc, 

Again, you are talking about behavior when a transient read error is
detected.  My focus is the due diligence done by the firmware during
write operations.

It is correct that one of the defining characteristics of nearline
vs. consumer drives is the retry behavior.  But that's not the point I
was trying to make.

What I was trying to convey was that:

1. Contrary to popular belief there is no inherent mechanical difference
   between consumer and nearline drives.  Same heads, arms, motors, etc.
   The premium you pay is not for "mechanical ruggedness".  That's what
   most people assume when they are charged more(*).

2. The difference is largely in how the firmware encodes stuff on the
   physical platters in the drive, the internal housekeeping overhead.
   That difference between consumer and nearline is getting bigger with
   each generation of drives.

That said, I'm also sure you can appreciate that media defect tolerances
are likely to be different between nearline and consumer kit despite
coming off the same assembly line.


(*) Seagate recently put out some SAS nearline drives that have a
different logic board than their SATA cousins.  So there's actually a
real hardware difference in that series.  The fatter PCB with dual
processors enables even better integrity protection (on par with "real"
enterprise drives) albeit at lower duty cycles.

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: upgrade advice

[Max Waterman]

2008/12/16 Redeeman <redeeman@metanurb.dk>:
> On Tue, 2008-12-16 at 06:44 -0500, Justin Piszcz wrote:
> Hehe, i myself am in a somewhat similar position.. i have a 6x300gb
> raid5 array.. And the disks appears perfectly fine through smart etc,
> but on the other hand, a SINGLE new disk today can present me with the
> same level of usable capacity very cheap, and ill bet using considerably
> less power.

...and noise (from fans/etc), which is probably more important for me
(it's next to the TV);

but a single disk doesn't provide any redundancy, which I would prefer
to have, I think.

> Thus i have chosen to very soon retire this setup, and make
> a replacement 8x1tb or 6x1.5tb raid6 for increased capacity and
> safety :)

Hrm. That's quite a Christmas present :) I don't think my sister can
stretch to that. I think I'd be lucky to get just a single one.

>
> i'd recommend you work towards deprecating LOTS of drives in favor of
> fewer bigger newer, and keep the old ones around as an archive/backup
> thing.

Yes, that sounds sensible. I guess I would like to do this in stages.

1) buy a 1TiB disk to do a backup
2) upgrade from 6+2/RAID5 to 7+1/RAID6
3) start building a new RAID6 starting with the 1TiB disk....somehow

Max.

Re: upgrade advice

[Max Waterman]

2008/12/16 Justin Piszcz <jpiszcz@lucidpixels.com>:
>
>
> On Tue, 16 Dec 2008, Max Waterman wrote:
>
>> Hi,
>>
>> I have a 8x200GB RAID5 array - 4 SATA and 4 EIDE. It has 2 spare, so
>> that's only 6 drives actually in the array, making 1TB of space (I
>> think that's right).
>>
>> The drives are quite old now, and people are asking what I want for
>> Christmas...so I'm wondering how to upgrade this setup. Clearly I
>> don't want to retire my existing drives prematurely, since they've
>> shown no sign of problem.
>>
>> My initial thought was to make a RAID1 out of a new 1TB disk and my
>> existing RAID5 array, but perhaps that doesn't make much sense...
>>
>> Perhaps I could get a 400GB drive (or more likely 500GB) and make a
>> RAID1 with a couple of the existing 200GBs?
>>
>> ...but I'm just guessing and the purpose of this email is to get advice.
>>
>> What do you guys think?
>>
>> Max.
>
> Tough call, depends on what you want to accomplish?  Availability or mass
> storage space?  You could continue running like you are now and use a 1TiB
> disk as your "scratch" area and rsync the important data off to your RAID-5
> nightly for example.

I think I want 'availability'. I would be really sick to lose the
contents, but it wouldn't be the end of the world...I have nowhere big
enough for a backup.

A 1TiB disk might be a good thing to do a backup while I convert from
RAID5 to RAID6, and be useful later in some way.

>
> Another option (probably not a good one) take two of your 200s and make a
> RAID1, leave your RAID-5 as-is, buy another 200GiB as a spare

I'm not sure what I'm ending up with here - 6x200/RAID5 + 2x200/RAID1
+ 200SPARE.

Seems a little messy to me...

> and then with
> the rest of the $ buy something else, there are no good alternatives, those
> are pretty old disks you got there, but I agree if its been running fine, no
> reason to replace it.

I guess my main reason is the fear that one will fail at some time
(soon); and while it rebuilds another will fail.

Seems like I should head for RAID6. I'm not sure I can even do that
making and restoring from a backup - can I?

Max.