How does Parity RAID work

How does parity work on a RAID 5 array?


I'm looking for a nice little RAID array for dedicated backups. I would love to have 2-4 TB of storage space as I have this nasty habit of digitizing everything. Therefore, I need a lot of memory and a lot of redundancy in the event of a drive failure. Also, I'll essentially be backing up the folders of 2-3 computers using one of the "Time Machine" clones for Linux. This array can be accessed over my local network via SSH.

I am having trouble understanding how RAID-5 achieves parity and how many drives are actually needed. One would assume it would take 5 drives, but I could be wrong. Most of the charts I've seen so far have just confused me. Apparently this is how RAID-5 works. Please correct me as I am sure I do not understand correctly:

It appears that drives 1-3 appear and function as a single massive drive () and the parity drives back up these drives. What strikes me as odd is that I typically see more than 3 storage drives on a diagram with only 1 or 2 parity drives. For example, let's say we have 4 1TB drives in a RAID 5 array, 3 running stores, and 1 running parity. We have 3TB of actual storage, but only 1TB of parity !?

I know I am missing something here, can someone help me? Which would be better for my application, RAID-5 or RAID-6? Fault tolerance is the top priority for me at this point as it is going to run over a network for personal use only, speed isn't enormously important.

Reply:


It simply XORs each corresponding bit from each drive. If you lose a drive, you can recreate the missing data.

To the background:

Assume that D is the XOR of the other columns. As long as you lose just one drive, you can find out what you lost.

Sometimes the stripe bit is spread across the drives, but the concept is the same.

For RAID-5, regardless of the number of drives, you only need 1 drive so that the parity is at least as large as the smallest drive in the array that you want to make RAID-capable.

RAID-5 for personal use is probably best because it is much less computationally than RAID-6.

RAID-6 is more complicated with Galois fields for calculating parity. And that can control parity calculations. However, you can lose more drives. However, if you are rebuilding your array as soon as a single failure occurs, stick with RAID-5.







I think here is a better diagram to show how parity works in RAID4 and RAID5

RAID4

RAID5



I would recommend reading this Wikipedia article on Raid 5 and Raid 6

http://en.wikipedia.org/wiki/Standard_RAID_levels#RAID_5_parity_handling

RAID 5 writes a parity block in each stripe, so that for stripe A of a 4-disk array, the parity check is written on the 4th disk with the data on disks 1, 2 and 3

For strip B, the parity block is on data carrier 3 with data on data carriers 1, 2 and 4, etc.

For example, if disk 4 fails, the data for strip B can be rebuilt because you know the data on disk 1 and 2 and run the parity check on disk 3.

If stripe B had a parity of "2" and data carrier 1 has data of "1" and data carrier 2 has data of "0", data carrier 4 must have data of "1" so that the data carrier with data = "1 "is described.

The entire hard drive can be rebuilt this way. RAID 6 extends this by creating 2 party blocks per stripe.

In terms of storage space for Raid 5, you will only ever lose one hard drive that has room for parity, as only parity blocks are written to per stripe. In raid 6 you will lose 2 disks, but you can also lose two disks instead of the one you can lose in raid 5;)

The Wikipedia article explains this better!


RAID 5 uses one drive for parity regardless of how many data drives are in the array. This means that the more drives that are added, the more efficient the available space becomes.

Parity is achieved by XORing the same block on each drive. The content of the parity drive is adjusted so that all drives set XOR to zero. This means that RAID 5 is limited by the smallest capacity of any of the drives in the array.

RAID 6 is similar, except that two simultaneous drive failures can be tolerated. This is useful because the process of resilienting an array from a single drive failure can be so strenuous that a second drive will fail.





If fault tolerance is your goal, RAID-6 provides enough redundancy to make two drives lose. Typically, RAID-5 will only tolerate a single drive failure.



We use cookies and other tracking technologies to improve your browsing experience on our website, to show you personalized content and targeted ads, to analyze our website traffic, and to understand where our visitors are coming from.

By continuing, you consent to our use of cookies and other tracking technologies and affirm you're at least 16 years old or have consent from a parent or guardian.

You can read details in our Cookie policy and Privacy policy.