Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I am a newbie in the world of storage and I am having a hard time digesting the physical architecture of some of the RAID levels. I am particularly interested in RAID 10, and 50. I asked the question specifically about RAID 10, because I feel if I understand that, I'll understand the other.

So, I get the definition of RAID 10 - "minimum 4 disks, a striped array whose segments are mirrored". If I've got 4 disks and Disks 1 and 2 are a mirrored pair, and Disks 3 and 4 are a mirrored pair - where does the data get striped?

Thanks.

share|improve this question
add comment

4 Answers

http://en.wikipedia.org/wiki/File:RAID_10.png

you have 2 raid 1 that go together into a raid 0 - raid 10.

look at this article as well - describes also the benefits etc. about it.

http://www.linux-mag.com/id/7928?hq_e=el&hq_m=1151565&hq_l=4&hq_v=bf05dd41dc

share|improve this answer
    
Thanks, I've read a number of articles, but for whatever reason, that linux-mag article seemed to help penetrate my denseness! –  Hank Jan 7 '11 at 14:48
    
I don't recommend it as a teaching aid, but flickr.com/photos/tirrell/128638167 is pertinent, and I find it rather funny. –  MadHatter Jan 7 '11 at 15:06
add comment

In the following A, B, C, D area pieces of data.

RAID 0:

DISK1    DISK2
  A        B

RAID 1:

DISK 1   DISK2
  A        A

RAID 10:

DISK 1   DISK2   DISK3   DISK4
  A        A       B       B

RAID 5: (p is parity, recovery information)

DISK 1   DISK2   DISKn
  A        B       p

RIAD 50: (p and q are parity, it's two of the above side by side...)

DISK 1   DISK2   DISKn   DISK4   DISK5   DISKm
  A        B       p       C       D       q

The Wikipedia RAID article has much more information and pretty pictures.

share|improve this answer
add comment

If disks 1 and 2 make a mirror A and disks 2 and 3 make a mirror B then the data would be striped across A and B. There are some good diagrams and articles here and here that help to explain it.

share|improve this answer
add comment

The data is striped at Raid0 first. Then that data is duplicated to each disk in the Raid1 array.

Say you have 16 physical disks for your Raid10 setup. This means you have 8 Raid1 arrays inside your Raid10. When you send data it will get split into 8 chunks. And each Raid1 array only needs to process one chunk of data.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.