Guide: Raid

RAID GUIDE

RAID- Redundant Array of Independent Disks

Raid is quite handy. It involves combining two or more hard drives to create either more space or more security. with certain arrays however, you can have both

There are 11 different arrays available presently.

Also, a few other terms need to be defined as well before I begin the listing of RAID arrays.



Fault Tolerance
- When one disk fucks up, so do the rest of them in your RAID array. The more backups you have of your information, the more fault tolerant you are.

Parity- The quality of being either odd or even. The fact that all numbers have a parity is commonly used in data communications to ensure the validity of data.

Ok, now we can start on the fun stuff, RAID J

RAID 0- Striped Disk Array without Fault Tolerance: 2 identical hard drives grouped together to make one big 1.

RAID 1- Mirroring and Duplexing: Provides disk mirroring. RAID 1 provides twice the read transaction rate of single disks and the same write transaction rate as single disks. 2 identical hard drives. 1 automatically backs up everything place on the first hard drive. Provides a small amount of fault tolerance.

RAID 2- Error-Correcting Coding: Not a common RAID array, RAID 2 stripes data at the bit level as opposed to the block level. RAID 2 is VERY expensive and requires MANY drives…

Either 10 data disks and 4 ECC disks, or 32 data disks and 7 ECC disks, which ends up being a lot of money…

RAID 3- Bit-Interleaved Parity: Provides byte-level striping with a dedicated parity disk. RAID 3, which cannot service simultaneous multiple requests, is also not utilized very often for this exact reason.

RAID 4- Dedicated Parity Drive: A common RAID configuration, RAID 4 uses block-level striping (like RAID 0) with a parity disk. If a data disk fails, the parity data is used to create a replacement disk. A disadvantage to RAID 4 is that the parity disk can create write bottlenecks.

RAID 5- Block Interleaved Distributed Parity: Uses data striping at the byte level and stripe error correction information. This produces excellent performance and good fault tolerance. RAID 5 is one of the most popular implementations of RAID.

RAID 6- Independent Data Disks with Double Parity: Uses block-level striping with parity data distributed across all disks.

RAID 7- A common configuration among Storage Computer Corporations that adds caching to RAID 3 or 4.

RAID 0+1- A Mirror of Stripes: 2 raid 0 configurations in raid 1

RAID 10- A Stripe of Mirrors: Not one of the original RAID levels, multiple RAID 1 mirrors are created, and a RAID 0 stripe is created over these. Quite similar to RAID 0+1



RAID S
- EMC Corporation's proprietary striped parity RAID system used in its Symmetrix storage systems.
 
name='maddcore' said:
Nice definitions, how about a set-up guide?

Setup varies by motherboard and raid controller.... Resort to your motherboards manual for a detailed setup guide or the users manual for your raid controller card.
 
name='maddcore' said:
Nice definitions, how about a set-up guide?

I guess it would depend what adapter you have.

I work with IBM ServeRAID adapters which support a large selection of RAID types. Easy to configure via the GUI...

Glen.
 
lo,

If i was to buy another HD exactly the same as my one atm, could I put it in RAID 0 without having to format? and use a software driver? I cba reinstalling windows again.
 
OCE - Online capacity expansion - Increasing the size of a raid array file system like raid 5 or 6 by adding a new hard disk to the array and having the filesystem automatically expand and restripe the raid array to include the new disk while never going down (data is still accessible just not as much transfer bandwidth)
 
Forgive me for being a bit stupid (omnes : perish the thought VB) but I had kinda grasped what the important (0,1,5) versions of RAID are. However, things like this "Uses data striping at the byte level and stripe error correction information." might as well be in Chinese for all I can make out.

Is data striping a good thing?

Is the byte level better than the block level?

As far as I can ascertain people use RAID 0 or 1 mostly. RAID 0 would be "one disk dies, lose all your data" and RAID 1 would be "buy two identical hard-drives to have the capacity of one of them".

So I'm completely at a loss as to why anyone would bother with all the hassle. Unless they were secretely the Sultan of Brunei with money to burn, the few extra ms saved per access is vital, or they had some extremely important information to back up and wanted to use a method other than the cheap optical medium.

Please, please put me right and explain why on earth it's worth such huge time, effort and expenditure. Is it really that much faster?
 
Raid 0 If one dies your erm fluffed. gives faster loading times than single drive

Raid 1 means that if one dies you still have your work. same speed I believe

It may be a good idea to have a simple version, so that people like me dont have to strain their brain.
 
Raid 5: 3+ drives, you have the capacity of [amount of drives]-[capacity of a single drive]. Drives need to be similar. When a drive dies you still have your data but you need to put in a new drive and update it with the rest. Also gives higher performance, although not nearly as much as raid0.

That's pretty much the complete simplified guide, together with above post :')
 
Just wondering what exactly is meant by 'identical hard drive'? Does that refer to a hard drive exactly the same as another, as in brand, capacity, interface etc...?

Sorry, but I'm a n00b
smile.gif
Specifically what I'm wondering is whether (for example) I can set up a RAID 0 with my SATA-3 WD Caviar Blue 250gb 16mb and my SATA-3 Seagate Barracuda 250gb 16mb and could I expect it to work as well as two identical brand drives with the same specs? On paper it seems like a silly question but I want optimal performance, same question regarding SSD's too I suppose
smile.gif
 
Back
Top