SATA RAID?

  • Thread starter Thread starter Tai Tze Hou \(Alvin\)
  • Start date Start date
T

Tai Tze Hou \(Alvin\)

Hi,

the specification on a motherboard for Serial ATA: RAID 0, RAID 1, RAID O+1,
Multiple RAID.

What does the different RAID stands for and what does it do?

Thanks.

Al
 
raid0 = striping. 1 segmant of data is writen to disk1, another to disk2 -
performance mode. however, if 1 HDD goes down, then all data is lost.
i`m lead to beleive that you can have more than 2 disks on this setup (the
silicon image sil3112 manual example is across 3 disks)

raid1 = mirroring. data is written to disc1, and also to disk2. there
IDENTICAL. this is stability/safety mode. if one HDD goes down, then the
raid controller will switch to disk2 and resume as normal with no data loss.

raid 0+1 = i`m not sure how it does it, but it's a combination of the above.
BOTH striping and mirroring on the same setup. for speed aswell as data
security.

multiple raid - not heard of. but i`m guessing it supports more than 1 raid
array.

i`m running SATA RAID0 array, and so far (3 weeks now) all is good. no data
corruption, loss, pauses or anything negative. just 71mb/s rather than the
30mb/s i was running before on an ATA133 maxtor :D

tim
 
Tai Tze Hou (Alvin) said:
Hi,

the specification on a motherboard for Serial ATA: RAID 0, RAID 1, RAID O+1,
Multiple RAID.

What does the different RAID stands for and what does it do?
Click to expand...

I'm not all that familiar with RAID arrays but a quick google for 'raid
configurations' I can only find reference to three types of RAID.

See: http://www.linkcomputers.com/uraidsconf.html for an example.

How many SATA controllers does your board support? I would have thought RAID
0+1 would be similar to RAID 5 but that requires at least three controllers
and three drives, say two 80GBs and one 160GB. It stripes to the 80s (in
this example) and mirrors to the 160 for both speed and data security.
 
raid 0+1 = i`m not sure how it does it, but it's a combination of the
above. BOTH striping and mirroring on the same setup. for speed
aswell as data security.
Click to expand...

ya it means 2 drives are striped, and another 2 are a mirror for the first
too

also, there is a botched standard raid 1.5 (or something like that)
by some company i can't rememeber, you get both mirroring and striping
using 2 drives, basicly same as a normal mirror, but when reading data,
instead of using the primary drive, it reads first bit from drive 1, then
bit 2 from drive 2, so as it get the striping bit

it's more complicated than that, but thats the basic idea, i forget which
company does makes em, the review i read wasn't that impressed really
 
Except that site uses the wrong definition for RAID. They say "Redundant
Array of *Inexpensive* Disks" when, in fact it should be "Independant". I
have written the web-master about it.
Click to expand...

The original was "Inexpensive", which it was when it was developed (they
were competing against mainframe arrays at the time). Was changed once the
prices caught up so that now Independant makes more sense.
 
JT said:
The original was "Inexpensive", which it was when it was developed (they
were competing against mainframe arrays at the time). Was changed once the
prices caught up so that now Independant makes more sense.
Click to expand...

You sure? Do you have corroborating evidence? I'm not being a smart-arse, I
really want to know. Just about every definition I've seen says
*Independant*, that was the point of it, all your data was shared between
disks, not all on the one.
 
What is RAID?

RAID (Redundant Array of Independent Disks) is an acronym first used in a 1988 paper by Berkeley researchers David Patterson, Garth
Gibson and Randy Katz. It described array configuration and applications for multiple inexpensive hard disks (aka Redundant Arrays
of Inexpensive Disks), providing fault tolerance (redundancy) and improved access rates.

either way will get your poiint accross
 
~misfit~ said:
You sure? Do you have corroborating evidence? I'm not being a smart-arse, I
really want to know. Just about every definition I've seen says
*Independant*, that was the point of it, all your data was shared between
disks, not all on the one.
Click to expand...

He's correct. RAID (Redundant Array of Inexpensive Disks) is a series of
storage technologies described in 1987 by three researchers from UC
Berkeley: David Patterson, Garth Gibson and Randy Katz. RAIDs were
conceived to overcome the data reliability problems associated with disk
storage subsystems while increasing various performance aspects.

However, another aspect of the idea was to use readily available and
'inexpensive' hard drives to accomplish it.

Abstract from their paper
http://sunsite.berkeley.edu/TechRepPages/CSD-87-391

A Case for Redundant Arrays of Inexpensive Disks (RAID)

David A. Patterson, Garth A. Gibson and Randy H. Katz
CSD-87-391

Increasing performance of CPUs and memories will be squandered if not
matched by a similar performance increase in IO. While the capacity of
Single Large Expensive Disk (SLED) has grown rapidly, the performance
improvements of SLED has been modest. Redundant Arrays of Inexpensive Disks
(RAID), based on the magnetic disk technology developed for personal
computers, offers an attractive alternative to SLED, promising improvements
of an order of magnitude in performance-reliability, power consumption, and
scalability. This paper introduces five levels of RAIDs, giving their
relative cost/performance, and compares RAIDs to an IBM 3380 and a Fugitsu
Super Eagle

------------------

One can be fairly confident that the originators of a term know what it
stands for ;)

"Independent disks" is a later 'interpretation' and, if you want to really
get confused, the RAID advisory board wants to rename it "intelligent disks."
 
~misfit~ said:
I'm not all that familiar with RAID arrays but a quick google for 'raid
configurations' I can only find reference to three types of RAID.

See: http://www.linkcomputers.com/uraidsconf.html for an example.

How many SATA controllers does your board support? I would have thought RAID
0+1 would be similar to RAID 5 but that requires at least three controllers
and three drives, say two 80GBs and one 160GB. It stripes to the 80s (in
this example) and mirrors to the 160 for both speed and data security.
Click to expand...

That's not how RAID 5 works. RAID 5 stripes all the (equal sized) drives
with data and error correction coding (e.g. parity). Data and the ECC are
stripe rotated among the three (can be more) drives to spread the wear and
tear equally among them.

Data can be recovered with any two of the three.

Conceptually it looks like

Disk 1 Disk 2 Disk 3

data1 data2 ecc
ecc data3 data4
data5 ecc data6
data7 data8 ecc
..
..
..
 
David Maynard said:
He's correct. RAID (Redundant Array of Inexpensive Disks) is a series of
storage technologies described in 1987 by three researchers from UC
Berkeley: David Patterson, Garth Gibson and Randy Katz. RAIDs were
conceived to overcome the data reliability problems associated with disk
storage subsystems while increasing various performance aspects.

However, another aspect of the idea was to use readily available and
'inexpensive' hard drives to accomplish it.

Abstract from their paper
http://sunsite.berkeley.edu/TechRepPages/CSD-87-391

A Case for Redundant Arrays of Inexpensive Disks (RAID)

David A. Patterson, Garth A. Gibson and Randy H. Katz
CSD-87-391

Increasing performance of CPUs and memories will be squandered if not
matched by a similar performance increase in IO. While the capacity of
Single Large Expensive Disk (SLED) has grown rapidly, the performance
improvements of SLED has been modest. Redundant Arrays of Inexpensive Disks
(RAID), based on the magnetic disk technology developed for personal
computers, offers an attractive alternative to SLED, promising improvements
of an order of magnitude in performance-reliability, power consumption, and
scalability. This paper introduces five levels of RAIDs, giving their
relative cost/performance, and compares RAIDs to an IBM 3380 and a Fugitsu
Super Eagle

------------------

One can be fairly confident that the originators of a term know what it
stands for ;)

"Independent disks" is a later 'interpretation' and, if you want to really
get confused, the RAID advisory board wants to rename it "intelligent
Click to expand...
disks."

Thanks for that David. I'm learning all the time.
 
David Maynard said:
That's not how RAID 5 works. RAID 5 stripes all the (equal sized) drives
with data and error correction coding (e.g. parity). Data and the ECC are
stripe rotated among the three (can be more) drives to spread the wear and
tear equally among them.

Data can be recovered with any two of the three.

Conceptually it looks like

Disk 1 Disk 2 Disk 3

data1 data2 ecc
ecc data3 data4
data5 ecc data6
data7 data8 ecc
Click to expand...

OK, thanks. I'm not au-fait with RAID, I just found that description of RAID
5 after a google search. The web page was obviously wrong.

Cheers,
 
it's DFI that do raid1.5.

tim

Geoff said:
ya it means 2 drives are striped, and another 2 are a mirror for the first
too

also, there is a botched standard raid 1.5 (or something like that)
by some company i can't rememeber, you get both mirroring and striping
using 2 drives, basicly same as a normal mirror, but when reading data,
instead of using the primary drive, it reads first bit from drive 1, then
bit 2 from drive 2, so as it get the striping bit

it's more complicated than that, but thats the basic idea, i forget which
company does makes em, the review i read wasn't that impressed really
Click to expand...
 
Back
Top