Sign in with
Sign up | Sign in
Your question

To RAID or not to RAID?

Last response: in Overclocking
Share
Anonymous
a b K Overclocking
May 26, 2005 9:38:50 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Yes it sounds like a newbie question, anyway...

Which is beter for performance, a single hard drive of say, 160 GB or
two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
drive (want to test the nF4 hack) and I see a 160 gig drive is about
$20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
same manufacturer, is a striped config of two drives faster than a
single drive equal in size to the other two? Wouldn't the single drive
naturally be faster? Just thought I'd ask, and thanks.

More about : raid raid

May 27, 2005 5:19:02 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

ah,,,

No


<theredpriest@altavista.com> wrote in message
news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.
>
Anonymous
a b K Overclocking
May 27, 2005 5:43:03 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

<theredpriest@altavista.com> wrote in message
news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.
>

Read a review on the 1st SATA 2 drive from Hitachi. It is a fast drive but
absolutely no noticeable performance increase over SATA 1. As a matter of
fact, hard drives can't even meet sata 1 potential.
Related resources
Anonymous
a b K Overclocking
May 27, 2005 1:54:03 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

theredpriest@altavista.com wrote:

> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.

If RAID 0 isn't faster what is the point of it? Why would anybody do it?
Anonymous
a b K Overclocking
May 27, 2005 2:46:55 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Never done RAID, so here's a newbie answer to the newbie question.
From what I read, there's not that much to be noticed in RW terms from RAID.
But it does introduce a substantial reliability issue.
Not just because you have twice the hardware to malfunction, but because
many boards have RAID implementation issues.
If your data's valuable to you, stick to unRAIDed IDE, 'cos the only time
RAID makes any perceptible difference is when loading programs/Windows.
An atrocious trade-off IMNSHO.

theredpriest@altavista.com wrote:
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and
> the same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single
> drive naturally be faster? Just thought I'd ask, and thanks.
Anonymous
a b K Overclocking
May 27, 2005 6:37:19 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

I ask a question, so why is it so difficult for anyone to understand:
how should I know whether two RAID'ed drives are faster than a single
larger drive UNLESS I ASK? RAID was built for not only speed but
mirroring abilities as well. I can see two striped drives being faster
than a single drive OF A DIFFERENT SIZE AND MANUFACTURER but how would
I know unless I've used a raid system in the past, and I haven't.

The point of my question was not whether SATA II has any noticeable
increase in transfer speeds over SATA I (which I already know it
doesn't, that is, for the time being) or why RAID was developed in the
first place. And one word answers, although nice and to the point, are
just as vague as saying nothing at all.

>From my POV it's like this: I ask a question hoping somebody out there
will answer who actually has had experience with RAID configurations
and knows whether 2 drives really are faster than one drive of equal
size or not. I get the jist of these answers (that in fact they are)
but no indication at all that anyone actually KNOWS. I appreciate your
replies but they only indicate opinion and conjecture, not knowledge,
and really don't answer anything for me.

Maybe I ask and expect too much, after all it is only Usenet, but what
the hey, it's a free country right. Happy Mem Day to one and all!
Anonymous
a b K Overclocking
May 28, 2005 2:03:46 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

On Fri, 27 May 2005 14:37:19 -0700, theredpriest wrote:

> I ask a question, so why is it so difficult for anyone to understand:
> how should I know whether two RAID'ed drives are faster than a single
> larger drive UNLESS I ASK? RAID was built for not only speed but
> mirroring abilities as well. I can see two striped drives being faster
> than a single drive OF A DIFFERENT SIZE AND MANUFACTURER but how would
> I know unless I've used a raid system in the past, and I haven't.

Raid 0 is faster than no raid. If you really want to know about raid you
need to do some reading. there's many different raid levels which do
different things. Personally, I'd never use raid 0 alone without a good
backup procedure. Now 0+1 would be nice, but requires 4 drives. Here's a
starting point.

http://www.webopedia.com/TERM/R/RAID.html

--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm
May 28, 2005 3:26:36 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

raid is definatly faster
better sustained transfers, ive only used a 2 drive array, i would never go
back to using a single drive again
2x120gig sata hitachi 7200rpm 8mb cache
as for RW performance windows/games feel much more quicker for me

ask anyone what there ideal hard drives would be and i bet they say 2x
raptors in raid
<theredpriest@altavista.com> wrote in message
news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.
>
Anonymous
a b K Overclocking
May 28, 2005 1:52:52 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

theredpriest@altavista.com wrote:
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config?

Depends on the drives. If you're comparing an identical model drive, say two
80GB 7200.7's vs a 160GB 7200.7 then the two 80's will be faster in most if
not all tests. I give myself wiggle-room as RAID0 has on average very
slightly higher seek times than a single non-RAID0. Very slightly being in
around 0.1ms or so in the worst real-world case, though you can devise
pathelogical access patterns that will make a RAID0 setup choke. In most
real-world situations, you don't need to worry.

This is because the 80GB drive might have two 40GB platters, and the 160GB
drive would just have twice this, so 4 platters. This is not always the case
(such as the 120GB WD1200JB's having larger platters than an 80GB WD800JB at
the same point in time) but it generally holds. However, if you are
comparing (say) two 80GB 7200.7's vs a 160GB 7200.8 (if it existed) then you
have a couple of differences. The 7's have 80GB per platter, but the 8's
have 100GB per platter. Also the actuator or firmware could have been
improved in the 8, improving seek times or TCQ performance. Unfortunately,
in this case the 8 is oftes SLOWER than the 7, but there is certainly a
possibility that a newer model drive can beat two older drives in RAID0 in a
number of situations (particularily those that are seek-time limited).

[...]

--
Michael Brown
www.emboss.co.nz : OOS/RSI software and more :) 
Add michael@ to emboss.co.nz ---+--- My inbox is always open
Anonymous
a b K Overclocking
May 28, 2005 2:36:16 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

I had 4 drives in raid 0 and got pissed off, it kept getting corrupt after a
month. I divided out the identical drives and have had no problem since.

<theredpriest@altavista.com> wrote in message
news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.
>
Anonymous
a b K Overclocking
May 28, 2005 4:38:24 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

"Mickey" <mickey@looney.com> wrote in message
news:119h3vcg9ia1h4d@corp.supernews.com...
>I had 4 drives in raid 0 and got pissed off, it kept getting corrupt after
>a month. I divided out the identical drives and have had no problem since.

What controller?


--
Ed Light

Smiley :-/
MS Smiley :-\

Send spam to the FTC at
uce@ftc.gov
Thanks, robots.
May 29, 2005 3:25:03 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Ashp <ashleypride@gmail.com> wrote in news:1117184044.492.0@nnrp-t71-
02.news.clara.net:

> theredpriest@altavista.com wrote:
>
>> Yes it sounds like a newbie question, anyway...
>>
>> Which is beter for performance, a single hard drive of say, 160 GB or
>> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
>> drive (want to test the nF4 hack) and I see a 160 gig drive is about
>> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and
the
>> same manufacturer, is a striped config of two drives faster than a
>> single drive equal in size to the other two? Wouldn't the single
drive
>> naturally be faster? Just thought I'd ask, and thanks.
>
> If RAID 0 isn't faster what is the point of it? Why would anybody do
it?
>

It is faster. 50% faster for me.
Anonymous
a b K Overclocking
May 30, 2005 5:19:20 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

<theredpriest@altavista.com> wrote in message
news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.
>

http://faq.storagereview.com/tiki-index.php?page=Single...

--
Derek
Anonymous
a b K Overclocking
May 30, 2005 11:32:03 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Mickey wrote:
> I had 4 drives in raid 0 and got pissed off, it kept getting corrupt
> after a month. I divided out the identical drives

You mean you de-RAIDed them ?

and have had no
> problem since.
> <theredpriest@altavista.com> wrote in message
> news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
>> Yes it sounds like a newbie question, anyway...
>>
>> Which is beter for performance, a single hard drive of say, 160 GB or
>> two 80 gig drives in a RAID 0 config? Going to get a SATA II hard
>> drive (want to test the nF4 hack) and I see a 160 gig drive is about
>> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and
>> the same manufacturer, is a striped config of two drives faster than
>> a single drive equal in size to the other two? Wouldn't the single
>> drive naturally be faster? Just thought I'd ask, and thanks.
Anonymous
a b K Overclocking
May 30, 2005 1:50:14 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

<theredpriest@altavista.com> wrote in message
news:1117154330.847395.175210@z14g2000cwz.googlegroups.com...
> Yes it sounds like a newbie question, anyway...
>
> Which is beter for performance, a single hard drive of say, 160 GB or
> two 80 gig drives in a RAID 0 config?

Better read and write speeds with RAID 0. But you double your failure rate.
If either drive fails, you lose all the data on both drives.

Your best bet for both speed and reliability is either RAID 0+1 or RAID 5.
Both give you the speed and increase, rather than decrease, your data
reliability. Unfortunately, RAID 0+1 doubles the drives needed (4 80GB's
gives you just 160GB in storage), and RAID 5 needs an expensive hardware
controller. If you're going to do RAID 0, then make sure you have another
drive in the system to store your data, and use the RAID for the OS and
programs.

BUT, the real question is, why do RAID 0 at all? Unless the rest of your
system is truly cutting edge, increasing the drive performance isn't really
going to have a major effect on overall system performance. So, your games
might load faster, but game play isn't going to be improved all that much.
Same with any other program.


Bill G






Going to get a SATA II hard
> drive (want to test the nF4 hack) and I see a 160 gig drive is about
> $20 cheaper than two 80G drives. Assuming a 7200rpm disc speed and the
> same manufacturer, is a striped config of two drives faster than a
> single drive equal in size to the other two? Wouldn't the single drive
> naturally be faster? Just thought I'd ask, and thanks.
>
Anonymous
a b K Overclocking
May 30, 2005 8:18:42 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

With all the talk (wherever RAID is mentioned) of RAID 0 being so dangerous
because if 1 drive fails they are both useless, I wonder if that matters so
much. If a single drive fails it's useless, unless you want to pay a company
to open it up and recover the data. Seems like keeping good backups is the
answer.

--
Ed Light

Smiley :-/
MS Smiley :-\

Send spam to the FTC at
uce@ftc.gov
Thanks, robots.
May 31, 2005 10:57:49 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Ed Light wrote:
> With all the talk (wherever RAID is mentioned) of RAID 0 being so dangerous
> because if 1 drive fails they are both useless, I wonder if that matters so
> much. If a single drive fails it's useless, unless you want to pay a company
> to open it up and recover the data. Seems like keeping good backups is the
> answer.


With two drives, there is twice the chance that one of them will fail
compared to a single drive.

But you're right, good backups are the answer!
Anonymous
a b K Overclocking
May 31, 2005 10:57:50 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

"andrew" <andrew0720@f2s.com> wrote

> With two drives, there is twice the chance that one of them will fail
> compared to a single drive.

I'm not much at math, but if, say, 2% of the drives in a certain line go
bad, then it's not much of a risk. Then, wouldn't it be just some small
percent more likely that one or both of a pair would be a 2%'er?


--
Ed Light

Smiley :-/
MS Smiley :-\

Send spam to the FTC at
uce@ftc.gov
Thanks, robots.
May 31, 2005 8:54:33 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Ed Light wrote:
> "andrew" <andrew0720@f2s.com> wrote
>
>
>>With two drives, there is twice the chance that one of them will fail
>>compared to a single drive.
>
>
> I'm not much at math, but if, say, 2% of the drives in a certain line go
> bad, then it's not much of a risk. Then, wouldn't it be just some small
> percent more likely that one or both of a pair would be a 2%'er?

Twice the number of drives so twice the chance one will fail.... so 4%.

We have a LOT of *BIG* RAID arrays in work (multi Terabytes each) and
I'd guess we are replacing drives at a rate of about one per week (they
are hot-swap arrays which auto-rebuild from the mirror). When you do
the math on the array, they are only failing at around the rate the
manufacturer quotes.

Think of it this way.... if you make a RAID array with 260 drives and
the manufacturer states a MTBF of say, 5 years..... one drive will fail
per week on average (5 years X 52 weeks = 260)!! Simple probability!
Anonymous
a b K Overclocking
May 31, 2005 9:00:18 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

"andrew" <andrew0720@f2s.com> wrote
> Think of it this way.... if you make a RAID array with 260 drives and the
> manufacturer states a MTBF of say, 5 years..... one drive will fail per
> week on average (5 years X 52 weeks = 260)!! Simple probability!

But not when the array was originally constructed, and all the drives were
new?


--
Ed Light

Smiley :-/
MS Smiley :-\

Send spam to the FTC at
uce@ftc.gov
Thanks, robots.
June 1, 2005 10:52:08 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Ed Light wrote:
> "andrew" <andrew0720@f2s.com> wrote
>
>>Think of it this way.... if you make a RAID array with 260 drives and the
>>manufacturer states a MTBF of say, 5 years..... one drive will fail per
>>week on average (5 years X 52 weeks = 260)!! Simple probability!
>
>
> But not when the array was originally constructed, and all the drives were
> new?


MTBF = ***MEAN*** time between failure..... A new drive could die
tomorrow or could last for 20 years.
Anonymous
a b K Overclocking
June 1, 2005 5:09:14 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Ed Light wrote:

> I'm not much at math, but if, say, 2% of the drives in a certain line go
> bad, then it's not much of a risk. Then, wouldn't it be just some small
> percent more likely that one or both of a pair would be a 2%'er?

The chance of a used drive going bad is 100%. The question of when they
bad is different.

Ash.
Anonymous
a b K Overclocking
June 1, 2005 5:40:37 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Ed Light wrote:


> But not when the array was originally constructed, and all the drives were
> new?

He said on *average* a harddrive would fail every week, basically at the
end of the 5 years most or all of the drives would (probably) be dead.
(The MTBF used here is very low compared to harddrives of today at 43,800).

However you are correct that at the start of the array the likely hood
of each indivual drive failing is slim. BUT The probablity of any drive
in the array failing gets higher.

ie. Lets say on day 1 the chance of a drive failing is 99.9%
Probability of a particular drive failing is 0.1% (1 in 1000 chance)
Probability of a random drive failing in the 260 2.6 % (1 in 38 chance)

Basically with a mirror you are doubling the chances of failure, if you
regulary back up, get good disks, and renew them regulary you may well
be safe, but you may well not be. Depends on how big a gambler you are.

Ash.
Anonymous
a b K Overclocking
June 2, 2005 4:40:24 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

andrew wrote:
[...]
> We have a LOT of *BIG* RAID arrays in work (multi Terabytes each) and
> I'd guess we are replacing drives at a rate of about one per week
> (they are hot-swap arrays which auto-rebuild from the mirror). When
> you do the math on the array, they are only failing at around the
> rate the manufacturer quotes.

I should point out that MTBF figures for hard drives are not quite what you
might expect at first glance. From below I see you have the right idea -
this post is for other people :)  Your average new SCSI drive has a MTBF of >
10^6 hours; the current Seagate Cheetahs have a 1400000 hour MTBF IIRC. Note
that 1.4 million hours is approximately 160 years! This doesn't mean at all
that you can run a batch of HDDs for a few hundred years until they all
fail, then calculate the average to be ~160 years. What is actually means is
that if you replace a drive at regular intervals (usually the warranty
length of the drive; 5 years for a standard SCSI drive) then you will only
have a drive fail on you (on average) every 160 years. Of course, this says
nothing about the failure rate of a single drive after it passes the 5 year
mark.

[...]

--
Michael Brown
www.emboss.co.nz : OOS/RSI software and more :) 
Add michael@ to emboss.co.nz ---+--- My inbox is always open
Anonymous
a b K Overclocking
June 3, 2005 10:34:22 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

I want to thank everyone who took the time to create a reply in this
thread, especially Derek and Ed for their links as well as Mike B. for
his in-depth replies. The storagereview.com link from Derek convinces
me I should go with a RAID config for performance, however the price of
Matxor and WD 250 gig drives right now is very tempting to me, like
buying three 80g drives for the price of two. As far as the risk of
losing data, well you're supposed to burn your critical data to CD or
DVD, right? But I thought I'd simply clone any data on SATA to IDE so
at least I'll have it around. I hope IDE will remain available and
cheap for a long, long time. Again fellahs, thaks for all your help.
Anonymous
a b K Overclocking
June 15, 2005 4:40:44 AM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

"andrew" <andrew0720@f2s.com> wrote in message
news:D 7i1bn$5fm$1@news.freedom2surf.net...
> >
> >
> Think of it this way.... if you make a RAID array with 260 drives and
> the manufacturer states a MTBF of say, 5 years..... one drive will fail
> per week on average (5 years X 52 weeks = 260)!! Simple probability!

Your math is incorrect here. The MEAN time is 5 years, so on average,
roughly half the drives will fail in less than 5 years, and half in more
than 5 years (technically, that would be the MEDIAN, but it's pretty close).
Your example has all of the drives failing in less than 5 years, so your
mean is around 2.5 years.

Bill G
!