Diskeeper 2008 & RAID 0

[Seidell23231]

Situation:
I am running a RAID 0 and have Diskeeper Pro Premier 2008 loaded and running
in the background.

My Question:
How important is it to Defrag and maintain a Defrag scheme in a RAID 0
environment?

Thanks ~~ Gunny

 

[Mike Brannigan]

Situation:
I am running a RAID 0 and have Diskeeper Pro Premier 2008 loaded and
running
in the background.

My Question:
How important is it to Defrag and maintain a Defrag scheme in a RAID 0
environment?

Thanks ~~ Gunny

Extremely as a RAID 0 is just a mirror set. So the disk you are reading from
is pretty much the same as a single disk and therefore subject to the same
degree of performance degradation due to defragmentation,

 

[Ken Blake, MVP]

Situation:
I am running a RAID 0 and have Diskeeper Pro Premier 2008 loaded and running
in the background.

My Question:
How important is it to Defrag and maintain a Defrag scheme in a RAID 0
environment?

RAID0 really doesn't affect the issue at all. The importance doesn't
change.

By the way, I used to run RAID0 here, but recently turned it off. I
perceive no decline in performance without it. But RAID0 does
substantially increase the risk of losing everything on your drive.
Without RAID 0, if one drive fails, you lose everything on that one
drive. But with RAID0, if either drive fails, you lose everything on
both drives.

For that reason, I think RAID0 is a bad bargain. You get substantially
increased risk in return for a very slight, or non-existent,
performance increase.

 

[Seidell23231]

Hey Mike,

Thanks for the info. Question - My RAID 0 is running on drives C & D. As
such do I setup Diskeeper for drive C or both C & D?

My Backup Plan consists of a Western Digital My Book Premium 750GB. I have
done a FULL backup and now doing incremental Backups daily.

Anything I'm missing? Suggestions?

Thanks ~~ Gunny

 

[DL]

Raid 0 is striping, Raid 1 is mirror

Extremely as a RAID 0 is just a mirror set. So the disk you are reading
from is pretty much the same as a single disk and therefore subject to the
same degree of performance degradation due to defragmentation,

 

[Seidell23231]

Hey Ken,

Thanks for your info. How long were you running a RAID 0 environment and
what experience occured that prompted your switch?

Again Thanks! Gunny

 

[Mike Brannigan]

Sorry - long day RAID 0 is a stripe set (with no redundancy). I would still
however go with defragmenting.

 

[Synapse Syndrome]

For that reason, I think RAID0 is a bad bargain. You get substantially
increased risk in return for a very slight, or non-existent,
performance increase.

I totally disagree with there being no performance increase. I have
benchmarked and rated with side-by-side tests, using the same hardware.
HDTach scores on my system were just as they supposed to be. Were you using
a PCI card? Were they IDE drives?

ss.

 

[Synapse Syndrome]

Thanks for your info. How long were you running a RAID 0 environment and
what experience occured that prompted your switch?

Again Thanks! Gunny

There are speed benefits, particularly if you are using very large files,
with sustained reads/writes. Using RAID-0 in video editing is where you
will really benefit from it. Access times and Burst speeds will not change
much at all, so the benefits on normal use are diminished.

I use it on my CAD workstation, for the OS partition and the Desktop User
Shell folder, and all files are kept off it. I've got twin WD Raptor
10,000rpm disks.

There is a lot of complicated juggling about you need to do when things go
wrong, even if you keep your data elsewhere, and as my setup is complicated
enough as it is, I will not be using RAID-0 again in my next build. The
benefits are not worth the hassle for me, personally.

ss.

 

[Kerry Brown]

I totally disagree with there being no performance increase. I have
benchmarked and rated with side-by-side tests, using the same hardware.
HDTach scores on my system were just as they supposed to be. Were you
using a PCI card? Were they IDE drives?

You can measure a difference in speed with RAID 0 vs. no RAID 0. If you will
actually perceive a difference depends on how you use the computer and the
RAID controller itself. Most RAID controllers built into the motherboard
don't have a dedicated CPU. They use the main CPU for calculations. This
means that if you are doing something that is CPU intensive that process may
suffer because the RAID driver will have priority for CPU time. When you are
running a benchmark program to test the RAID speed try watching a video at
the same time. Now run a couple of vm's and see what happens. In most cases
as you exercise the RAID array the other programs will suffer. Even though
the drive throughput may be slightly better most people would perceive
computer performance as decreased because the part they see appears to be
performing worse. The other thing to factor in is what most people do with
their computer. In the vast majority of cases the main slowdown from the
hard drive is actually moving the heads to where the data is. Very few
combinations of motherboard RAID controllers and consumer drives would
support real command queuing. This means that the bottleneck will still be
drive head positioning. RAID 0 may have a place but it is not for the casual
user with normal consumer grade computer equipment.

 

[NoStop]

There are speed benefits, particularly if you are using very large files,
with sustained reads/writes. Using RAID-0 in video editing is where you
will really benefit from it. Access times and Burst speeds will not
change much at all, so the benefits on normal use are diminished.

I use it on my CAD workstation, for the OS partition and the Desktop User
Shell folder, and all files are kept off it. I've got twin WD Raptor
10,000rpm disks.

There is a lot of complicated juggling about you need to do when things go
wrong, even if you keep your data elsewhere, and as my setup is
complicated
enough as it is, I will not be using RAID-0 again in my next build. The
benefits are not worth the hassle for me, personally.

ss.

With that FakeRAID controller, are you sure you can carry over those striped
drives to a new computer and they'll work?

Cheers.

--
What does Bill Gates use?
http://tinyurl.com/2zxhdl

Proprietary Software: a 20th Century software business model.

Be Afraid ... Be Very Afraid ... of Francis' RELATIVES!

Frank, hard at work on his Vista computer all day:
http://redwing.hutman.net/~mreed/warriorshtm/compost.htm

 

[Synapse Syndrome]

With that FakeRAID controller, are you sure you can carry over those
striped
drives to a new computer and they'll work?

I only use motherboards with Intel chipsets, so yes, they should. It's not
an issue anyway, as I only use RAID-0 for the OS and Desktop folder.

ss.

 

[Synapse Syndrome]

I only use motherboards with Intel chipsets, so yes, they should. It's
not an issue anyway, as I only use RAID-0 for the OS and Desktop folder.

Program Files is also on the same partition as the OS.

ss.

 

[Synapse Syndrome]

You can measure a difference in speed with RAID 0 vs. no RAID 0. If you
will actually perceive a difference depends on how you use the computer
and the RAID controller itself. Most RAID controllers built into the
motherboard don't have a dedicated CPU. They use the main CPU for
calculations. This means that if you are doing something that is CPU
intensive that process may suffer because the RAID driver will have
priority for CPU time. When you are running a benchmark program to test
the RAID speed try watching a video at the same time. Now run a couple of
vm's and see what happens.

Yes, I know this. Although in real-world use, this hardly ever happens.
When it does, it is a small percentage of the time I am using the computer,
and I do a lot of multitasking, with twin screens and large applications.

In most cases as you exercise the RAID array the other programs will
suffer. Even though the drive throughput may be slightly better most people
would perceive computer performance as decreased because the part they see
appears to be performing worse. The other thing to factor in is what most
people do with their computer. In the vast majority of cases the main
slowdown from the hard drive is actually moving the heads to where the data
is. Very few combinations of motherboard RAID controllers and consumer
drives would support real command queuing. This means that the bottleneck
will still be drive head positioning. RAID 0 may have a place but it is not
for the casual user with normal consumer grade computer equipment.

I don't disagree.

ss.

 

[Ken Blake, MVP]

Hey Ken,

Thanks for your info.

You're welcome. Glad to help.

How long were you running a RAID 0 environment and

2-3 years

what experience occured that prompted your switch?

I recently had an error that caused me to reload everything on the
drive from a backup. I took advantage of the opportunity to do what I
had wanted to do for a while (turning off RAID0), because I was
convinced that I was incurring extra risk for very little benefit.

 

[Ken Blake, MVP]

I totally disagree with there being no performance increase. I have
benchmarked and rated with side-by-side tests, using the same hardware.
HDTach scores on my system were just as they supposed to be. Were you using
a PCI card? Were they IDE drives?

No. They are SATA drives.

I haven't done any benchmarks, and I am perfectly willing to accept
that benchmarks will show a speed increase. However my experience in
day-to-use is that the increase was so small as to be imperceptible.

 

[Seidell23231]

What GREAT interaction this topic brings.

BACKGROUND ~~ I have 2 identical drives, both Seagate 360GB. When I setup
the RAID yesterday, I had both drives labeled and the second drive formated,
but empty. After the RAID had completed setting up, I noted that the second
hard drive changed from formatted with a label to "UNALLOCATED". The RAID
indicates the total space available at 670. Which would indicate it is
including the second UNALLOCATED hard drive. The RAID Controller and Intel
Matrix Storage Console, both show both HD's and the total available space at
670. The Computer Management / Disk Management, show the second drive, but
UNALLOCATED. The only difference is the My Computer area, wherein the
UNALLOCATED drive is not showing and it is not showing its number anywhere.

QUESTION ~~ Is the second drive's UNALLOCATED designation correct? Why did
it change from Formated and labeled to UNALLOCATED? As always, Thanks for
all the help! Gunny

 

[Seidell23231]

I know there are camps on both sides of the fence, so I come to both with
this question:

I will be using my HP Pavillion Elite m9150f, Intel 2 Core Quad QC6600, 3GB
PC-5300, 2 HD Seagate 360GB @ 7200, to do daily work, which means a lot of
time in MS Office 2007, specifically Excel and Access. I am in the process
of writing a book, so Word will be used in a large scale. Of course daily
backups will be the norm. I will be surfing the internet of course and
finally I will be spending 3-4 hours a day in Elder Scrolls: Oblivion and
Shivering Isles with a bried visit to Neverwinter Nights 2, Mask of the
Betrayor. For those not into the gaming worls, let me give a bried
description. They are all Stand Alone Role Playing Games. Yes they can also
be played over the internet, but I just don't enjoy it as much in that venue.
They are HEAVY duty games which require the top dog CPU's and Video CPUs as
well. So, with that being the consumption of daily use, will the RAID 0 be
of benefit or not. please provide a percentage, such as 80% benefit, 20% no
benefit.

As always, THANKS! Gunny

 

[Kerry Brown]

For what you describe you will not see a perceivable difference with or
without RAID 0. Non of the tasks you describe are disk intensive enough.
Even if you purchased a proper RAID controller and used 10,000 RPM drives
designed for RAID use you wouldn't see any noticeable difference.

 

[Seidell23231]

Thanks Kerry! What are your thoughts regarding the UNALLOCATED HD on my 2nd
to last post???

Thanks! Gunny