IDE RAID

[Tim]

Let me see, who would I believe?

Intel?

Someone that calls people wacko with no evidence?

Tough choice Ron.

- Tim

 

[Ron Reaugh]

Let me see, who would I believe?

Intel?

Intel agrees with me and not you wacko.

Someone that calls people wacko with no evidence?

With substantial evidence that's already been cited. Anyone can read the
whole thread and see for themselves.

Tough choice Ron.

Yep, you're outclassed...give it up.

 

[Leythos]

Nothing in that spec sheet says anything about HW RAID. The ICH5R does NOT
do HW RAID. There is no uP and NO buffer. Such are requirements of HW
RAID.

I may be missing something here, but I don't see anything that states
for something to be a raid controller that it must have a CPU or Cache.
As long as the chipset, using any firmware, handles the communications
with the drive and provides RAID 0/1/5 ability, it's Hardware Based
RAID.

If I use a non-RAID chipset and require the OS to process everything,
then it's soft RAID.

If I'm lucky enough to get cache and a CPU on the controller then I get
even faster RAID.

As I see it, the Promise or Intel RAID chipsets on some motherboards
provide the functionality needed to be considered hardware RAID. Sure,
they don't have their own CPU's, but they do have their own BIOS, do
have their own firmware, do take commands from the OS, and do allow the
creation, building, rebuilding of RAID Arrays before the OS is even
installed on the system.

 

[formerprof]

These ad hominem arguments -- like "wacko" and "clueless" don't strengthen
the case. Above all they don't clarify the disagreement if there is one.

 

[Paul]

I may be missing something here, but I don't see anything that states
for something to be a raid controller that it must have a CPU or Cache.
As long as the chipset, using any firmware, handles the communications
with the drive and provides RAID 0/1/5 ability, it's Hardware Based
RAID.

If I use a non-RAID chipset and require the OS to process everything,
then it's soft RAID.

If I'm lucky enough to get cache and a CPU on the controller then I get
even faster RAID.

As I see it, the Promise or Intel RAID chipsets on some motherboards
provide the functionality needed to be considered hardware RAID. Sure,
they don't have their own CPU's, but they do have their own BIOS, do
have their own firmware, do take commands from the OS, and do allow the
creation, building, rebuilding of RAID Arrays before the OS is even
installed on the system.

My understanding of hardware raid is as follows:

For a mirror:

If the OS prepares precisely one block of memory, with data
to be read or written, and the hardware solution takes that
block of memory and reads or writes to two disks, and only
returns "complete" status to the OS when both disks finish,
that is hardware RAID. If the OS has to issue two commands
to the hardware, saying write this to disk 0, then says write
this to disk 1, that is software RAID.

For a stripe:

If the OS prepares precisely one block of memory, and
issues one command to the hardware, and the hardware
alternates writing stripe-sized chunks of data to the
two drives, that is hardware RAID. If the OS chunkifies
the original large memory block, and alternates commands
to the two channels to write a stripe of data to the drives,
that is software RAID.

The difference is in the overhead. With DMA transfer, the
largest overhead of data movement by the processor is
removed. So, these days, it will be harder to tell whether
the solution is hardware or software based underneath. It
will be hard to tell from the remaining level of overhead,
to what extent the hardware hides the details of how the
disk subsystem is wired up, and what it is
(mirror or stripe).

The quality of any solution will be measured in two
parameter - max steady state bandwidth (HDTach) and
percent CPU while doing it. So, experiment and find out.

Whether RAID is hardware or software is based on the
level of abstraction. If the OS/driver, when viewing
the hardware, thinks it is dealing with a single disk,
when in fact the controller handles all the details of
running the RAID, that is a hardware controller. If the
OS is aware of the details underneath, to achieve basic
data transport, then it is a software based controller.

Notice that my definition doesn't cover implementation,
so you don't need to know the private details of how it
is done, to have a definition.

Just my two cents,
Paul

 

[Leythos]

Whether RAID is hardware or software is based on the
level of abstraction. If the OS/driver, when viewing
the hardware, thinks it is dealing with a single disk,
when in fact the controller handles all the details of
running the RAID, that is a hardware controller. If the
OS is aware of the details underneath, to achieve basic
data transport, then it is a software based controller.

Notice that my definition doesn't cover implementation,
so you don't need to know the private details of how it
is done, to have a definition.

So, if the idea that the OS thinks it dealing with 1 drive leads one to
assume it's a real hardware raid solution, then the Promise RAID 0/1
controller on the "ASUS PC-DL Deluxe" board is a hardware raid solution.
The OS doesn't see it as more than a single drive.

 

[Ron Reaugh]

I may be missing something here, but I don't see anything that states
for something to be a raid controller that it must have a CPU or Cache.

"RAID controller" is a generic term that could mean most anything. The
issue at hand is the precise and specific meaning of "hardware RAID".

I have already provide the specific criteria required to qualify as
"hardware RAID" earlier in this thread.

As long as the chipset, using any firmware, handles the communications
with the drive and provides RAID 0/1/5 ability, it's Hardware Based
RAID.

That is flat FALSE!

If I use a non-RAID chipset and require the OS to process everything,
then it's soft RAID.
Nope.

If I'm lucky enough to get cache and a CPU on the controller then I get
even faster RAID.

Nope, then you have true hardware RAID.

As I see it, the Promise or Intel RAID chipsets on some motherboards
provide the functionality needed to be considered hardware RAID.

That's flat FALSE.

Sure,
they don't have their own CPU's, but they do have their own BIOS, do
have their own firmware, do take commands from the OS,

Nope, during OS operation ALL the RAID functionality is handled by OS
device drivers and the firmware does nothing. The firmware is only used
during boot an pre OS activity.

and do allow the
creation, building, rebuilding of RAID Arrays before the OS is even
installed on the system.

Exactly...."before".

 

[Ron Reaugh]

These ad hominem arguments -- like "wacko" and "clueless" don't strengthen
the case. Above all they don't clarify the disagreement if there is one.

There is no case at issue. The definition of "hardware RAID" is well
established in the industry. There are just a few local wackos trying to be
revisionists. Folks engaged in such deserve to be dealt with accordingly.

 

[Ron Reaugh]

My understanding of hardware raid is as follows:

For a mirror:

If the OS prepares precisely one block of memory, with data
to be read or written, and the hardware solution takes that
block of memory and reads or writes to two disks, and only
returns "complete" status to the OS when both disks finish,

Exactly. And no host based x86 code nor host I/O bus structure is involved
in between. The block gets busmastered DMA-ed just once.

that is hardware RAID. If the OS has to issue two commands
to the hardware, saying write this to disk 0, then says write
this to disk 1, that is software RAID.

Exactly...that's what happens for both the ICH5R and onmobo Promise but not
the 3Ware nor other true HW RAID.

For a stripe:

If the OS prepares precisely one block of memory, and
issues one command to the hardware, and the hardware
alternates writing stripe-sized chunks of data to the
two drives, that is hardware RAID.
Exactly.

If the OS chunkifies
the original large memory block, and alternates commands
to the two channels to write a stripe of data to the drives,
that is software RAID.

Exactly and is what the ICH5R and Promise do.

The difference is in the overhead.

Right and for RAID 0 and RAID 1 that SW RAID overhead isn't that big.

With DMA transfer, the
largest overhead of data movement by the processor is
removed.

Right, but DMA does use bus bandwidth which can slow CPU and other I/O
operations.

So, these days, it will be harder to tell whether
the solution is hardware or software based underneath.

The first order way to tell is the presence of a uP on/in the RAID
controller that hosts a full low level full robust disk driver and that's
not x86 code most often. Also there'll be significant buffering onboard
that RAID controller where the data is held while the RAID processes operate
as you describe above.

It
will be hard to tell from the remaining level of overhead,
to what extent the hardware hides the details of how the
disk subsystem is wired up, and what it is
(mirror or stripe).

Exactly, the difference between SW/firmware RAID and hardware RAID for
simple RAID 0 and RAID 1 is SMALL. For RAID 10 and especially RAID 5 then
hardware RAID becomes distinctly superior.

What you didn't describe above is RAID 5. RAID 5 can be thought of as RAID
0 plus parity. True hardware RAID 5 uses its onboard uP to calculate the
parity using code hosted in the controller memory and nothing in the x86
host. That onboard uP access the data from its onboard cache thus adding NO
additional host bus structure overhead.

The quality of any solution will be measured in two
parameter - max steady state bandwidth (HDTach) and
percent CPU while doing it. So, experiment and find out.
Right.

Whether RAID is hardware or software is based on the
level of abstraction. If the OS/driver, when viewing
the hardware, thinks it is dealing with a single disk,
when in fact the controller handles all the details of
running the RAID, that is a hardware controller.
Yes.

If the
OS is aware

You mean any x86 host based code or driver involved in regular I/O to the
array.

of the details underneath, to achieve basic
data transport, then it is a software based controller.
Right.

Notice that my definition doesn't cover implementation,
so you don't need to know the private details of how it
is done, to have a definition.

Right but the reality of the real world provides some rather good litmus
tests to differentiate hardware from software/firmware RAID.

 

[Ron Reaugh]

So, if the idea that the OS thinks it dealing with 1 drive leads one to
assume it's a real hardware raid solution, then the Promise RAID 0/1
controller on the "ASUS PC-DL Deluxe" board is a hardware raid solution.

NO! The Promise has an OS level driver that knows it's talking to different
disks. That means the Promise is specifically a SW/firmware RAID.

The OS doesn't see it as more than a single drive.

Wrong, the OS at the Promise x86 basedf driver level does see it as more
than one disk and that OS level Promise driver in RAID 1 does do two writes.
Re-read Paul's or my earlier posts. The two x86 based data transfer and
writes is the critical issue.

 

[Ron Reaugh]

Nope.

Correction: YES!

 

[Leythos]

NO! The Promise has an OS level driver that knows it's talking to different
disks. That means the Promise is specifically a SW/firmware RAID.


Wrong, the OS at the Promise x86 basedf driver level does see it as more
than one disk and that OS level Promise driver in RAID 1 does do two writes.
Re-read Paul's or my earlier posts. The two x86 based data transfer and
writes is the critical issue.

I would love to know where you've determined, in a Promise document,
that the controller used on the PC-DL requires the OS to handle all the
data to both drives instead of sending it to the controller and the
controller handling the actual communications with the drives. If you
can post a link to the Promise document or some source other than you
own typing I would appreciate it.

 

[Tim]

Try a google search on "Ron Reaugh", peruse the results and check that the
posts on the numerous news groups are from (e-mail address removed).

So Ron, you write device drivers do you? Seems to me your a beginner in the
industry.

- Tim

 

[Ron Reaugh]

I would love to know where you've determined, in a Promise document,
that the controller used on the PC-DL requires the OS to handle all the
data to both drives instead of sending it to the controller and the
controller handling the actual communications with the drives. If you
can post a link to the Promise document or some source other than you
own typing I would appreciate it.

Re-read the whole thread. There is no place in that Promise chip nor in the
ICH5R for all the things that would have to be there for it to be true HW
RAID.

http://www.3ware.com/products/pdf/AMCC_MPG_0519.pdf
http://graphics.adaptec.com/pdfs/raid_soft_v_hard.pdf

http://nas.darma.com/support/hcl.html
Note the column where "not true HW RAID" is shown vs various RAID products.
Both Promise and the ICH5R get "not true HW RAID".

 

[Ron Reaugh]

Try a google search on "Ron Reaugh", peruse the results and check that the
posts on the numerous news groups are from (e-mail address removed).

A clear technique of a troll and stalker. When you can't carry the day
technically then attack the person.

You've been outclassed and are a FRAUD!

 

[Ron Reaugh]

Ron, where was the personal attack?
You have been given a lot of opportunity to provide evidence for your
arguments and have not produced any.

You are a bald faced liar as ayone can see.

 

[Tim]

Ron, where was the personal attack?
You have been given a lot of opportunity to provide evidence for your
arguments and have not produced any.

Since you claimed to be a device driver writer, my thinking was that perhaps
in prior posts elsewhere you may actually have furnished evidence and
information to support this. In the few I looked at (there were 10 pages of
results), I saw nothing to support any of your arguments or claims. I have
also researched your side of the argument and have nothing that I could add
or give to your to support what you are claiming. Is that balanced? Is it
fair?

Me, a fraud? One thing I have learnt in life is that when a person makes a
baseless accusation such as this, is that they are generally talking about
themselves, either that or they are being irrational as they have some other
major problem. I am not too sure why you insist on calling people names and
not providing any support for your claims... You do have the google toolbar
don't you?

The really big irony here is that there is the remote possibility that you
*could* have been right. However you have lost all opportunity at gaining
credit for exposing a Major Deceipt. You can flit around news groups for the
rest of your life if you wish, or you can learn to argue your case clearly
and cleanly and be of great benefit to all and gain something for yourself.
If you need an example of how to do that, then take a look at Paul's work.

It does annoy me when I see responses come from people such as yourself that
claim to be authoritative, but are wrong. Novices have a hard enough job
digesting what is in front of them without deliberately misleading
information. Calling people Wrong or Wacko is deliberately misleading.

BTW: I suggest you check with Adaptec about your definition of RAID. Adaptec
has made many RAID cards in the past that have not had cache memory as
standard.

Also, how do you know that the 82801 chip does not contain a processor? If
it doesn't contain a processor of some sort then what is the firmware for?
It seems you don't know what firmware is. If Intel wants to put a processor
of some time inside such a chip and not tell anyone then thats their
business. If Intel has not and has implemented their RAID purely in hardware
then that is there business.

Paul's definition of RAID is solid IMHO.

So, again Ron, come clean. How about some evidence? Show us some evidence
that supports your assertion that the OS has to issue (via the supplied
driver) multiple IO's when the application thinks it is dealing with 1
physical volume? Try Windows Performance Monitor for a start and write a
test app with IO's above and below the stripe size and look for a
correlation between reported physical IO's and logical IO's taking into
account that not all IO's are actually IO's. Since you have HDD ddk
experience, then surely you can write a layered device driver to intercept
the IO's from Intel driver and the physical device? I know this can be done
as this is how one can implement device level encryption - PGP Disk does /
used to do it. You could implement a custom Perfmon object to show counts of
reads and writes etc. If you need the source code for this I can help your
out.

- Tim

 

[Leythos]

Re-read the whole thread. There is no place in that Promise chip nor in the
ICH5R for all the things that would have to be there for it to be true HW
RAID.

http://www.3ware.com/products/pdf/AMCC_MPG_0519.pdf

Interesting "All 3ware products incorporate an onboard processor for
true hardware RAID performance." Notice, they only mention their product
as factoring performance, not that it's anything more than any other
RAID (except for performance).

They do say this "Software RAID schemes use the system processor, occupy
host memory, and consume CPU cycles." But they don't say what
constitutes "software raid" - from their own documents it appears that
they describe software RAID as the standard OS RAID, having nothing to
do with onboard RAID controllers. The firmware and chipset interface for
the Promise RAID controller on the PC-DL Deluxe clearly meets the
hardware requirement as based in this document. I would also imagine
that the Promise SX-6000 ATA 6 channel RAID card (which you said didn't)
clearly is hardware RAID, even in your mind it would be, if we used your
logic.

http://graphics.adaptec.com/pdfs/raid_soft_v_hard.pdf

This document clearly makes NO distinction between onboard RAID
controllers and other boarded RAID controllers. In fact, it only talks
about a RAID CARD and a SCSI non-RAID card in testing. If you look at
the document closely it does talk about dedicated processors and
firmware, but it clearly separates software RAID as being something that
is not based around a chipset, the talk about it being OS driven. The
firmware and chipset interface for the Promise RAID controller on the
PC-DL Deluxe clearly meets the hardware requirement as based in this
document. I would also imagine that the Promise SX-6000 ATA 6 channel
RAID card (which you said didn't) clearly is hardware RAID, even in your
mind it would be, if we used your logic.

http://nas.darma.com/support/hcl.html
Note the column where "not true HW RAID" is shown vs various RAID products.
Both Promise and the ICH5R get "not true HW RAID".

So, it looks like a vendor is saying that the Promise is not "true"
hardware RAID, but I don't see anything other than their linux variant
that suggest where it doesn't meet the RAID spec, in fact, it only
suggests that you use a different driver. I don't believe that this
vendor is credible in this discussion.

 

[Ron Reaugh]

Interesting "All 3ware products incorporate an onboard processor for
true hardware RAID performance." Notice, they only mention their product
as factoring performance, not that it's anything more than any other
RAID (except for performance).

They do say this "Software RAID schemes use the system processor, occupy
host memory, and consume CPU cycles." But they don't say what
constitutes "software raid" - from their own documents it appears that
they describe software RAID as the standard OS RAID, having nothing to
do with onboard RAID controllers. The firmware and chipset interface for
the Promise RAID controller on the PC-DL Deluxe clearly meets the
hardware requirement as based in this document.

That's flat FALSE! From your own quote above "Software RAID schemes use the
system processor, occupy host memory, and consume CPU cycles." The
Promise and ICH5R both do all those things..ergo SW/firmware RAID.

I would also imagine
that the Promise SX-6000 ATA 6 channel RAID card (which you said didn't)

Nonsense. What I said was "don't know" and I went on to say that some
Promise cards are semi-HW RAID.

clearly is hardware RAID, even in your mind it would be, if we used your
logic.
Clueless.


This document clearly makes NO distinction between onboard RAID
controllers and other boarded RAID controllers.

HUH? The issue at hand is SW vs HW RAID and NOT overboard where you've
jumped.

In fact, it only talks
about a RAID CARD and a SCSI non-RAID card in testing. If you look at
the document closely it does talk about dedicated processors and
firmware, but it clearly separates software RAID as being something that
is not based around a chipset, the talk about it being OS driven. The
firmware and chipset interface for the Promise RAID controller on the
PC-DL Deluxe clearly meets the hardware requirement as based in this
document.

Pure CARP as anyone who reads it can see for themselves.

I would also imagine that the Promise SX-6000 ATA 6 channel
RAID card (which you said didn't) clearly is hardware RAID, even in your
mind it would be, if we used your logic.
Clueless.


So, it looks like a vendor is saying that the Promise is not "true"
hardware RAID, but I don't see anything other than their linux variant
that suggest where it doesn't meet the RAID spec, in fact, it only
suggests that you use a different driver. I don't believe that this
vendor is credible in this discussion.

Wacko.

 

[Leythos]

Clueless.

Nice response Ron, I've added you to the list of people I place no
credit with (like some of the others here have done with you).

It will be interesting when they remove all the device stubs from an OS
for you so that you can get real hardware RAID as you think it is.