Alright comp nerds.

[phunky.buddha]

Quoted post[/post]]

Quoted post[/post]]
thats why if i was a home user, i would have 3 drives in raid 0+1. but not even i cant afford that so, frequent back is a must. we all know windoes; even mr. gates himself has problems with it.

id still take unregistered because my data isnt all the important, if it is, usually have 2 back ups of it on dvd+rs and the cost out weighs the benefits for home use.

Screw 0+1, with three drives, you want RAID 5.

Especially because 0+1 is impossible with 3 drives. You need an even number, and you need at least 4.

 

[reikoshea]

lol thats what i thought, but i figured id just wait for someone else to say it.

 

[endlesszeal]

oh haha my bad. i was thinking either 3 or 4. anyway, i read a little on raid 5 and still a little confused. so puts a little chuck of the data on to 3 or more hard drives?

what are the advantages of it?

 

[reikoshea]

really really big single directories.

fast read speeds, but kinda slow write speeds.

still has parity and data protection like the other raids

 

[Sabz5150]

Quoted post[/post]]
oh haha my bad. i was thinking either 3 or 4. anyway, i read a little on raid 5 and still a little confused. so puts a little chuck of the data on to 3 or more hard drives?

what are the advantages of it?

You can rip one of the drives out while it's running and you'd never notice.

RAID 0 merely stripes the data across multiple drives. Take two 400GB drives, RAID 'em and you've got a single 800GB drive.

RAID 1 is mirroring. Two 400GB drives will still be a 400GB volume, but you will have a perfect backup should one drive fail.

RAID 0+1 is a combination of the two aforementioned methods. Four 400GB drives would give you a single 800GB volume with a second 800GB volume as a mirror should any drive fail.

RAID 5 uses a checksum drive. Each stripe writes its checksum data to that drive. Five 400GB drives would give you a 1.2TB volume with the fifth drive being used for the checksums. Basically, if one drive fails, the array can reconstruct the data using the checksums. If you're working with important data and/or a volume of that size, RAID 5 is the only way to go.

Now, RAID 6 is similar to RAID 5 but it uses TWO checksum drives. Same concept, except this is fault tolerant up to two drives failing. Once again, uber important data and/or massive volumes.

 

[phunky.buddha]

Quoted post[/post]]
really really big single directories.

fast read speeds, but kinda slow write speeds.

still has parity and data protection like the other raids

Nope, fast write speeds too. You're always writing to multiple drives.

Quoted post[/post]]
RAID 5 uses a checksum drive. Each stripe writes its checksum data to that drive. Five 400GB drives would give you a 1.2TB volume with the fifth drive being used for the checksums. Basically, if one drive fails, the array can reconstruct the data using the checksums. If you're working with important data and/or a volume of that size, RAID 5 is the only way to go.

Now, RAID 6 is similar to RAID 5 but it uses TWO checksum drives. Same concept, except this is fault tolerant up to two drives failing. Once again, uber important data and/or massive volumes.

Sort of... but you have to mention that 5 and 6 use distributed parity. You don't have a dedicated parity drive- it's split up among all the other drives in a way that you have one drive's worth of data used as parity on all the drives. If you have a RAID 5 setup with 5x 100GB drives, you have 20GB of parity information on each one, giving you a total of 400GB of storage with 100GB of total parity information. None of the drives is a dedicated parity drive.