• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 395
  • Last Modified:

RAID 6

So the r510 dell server has a 8 drive chassis. Currently there are total of 6 drives, (2) 250GB for OS set to RAID 1 and (4) 1TB drives for data set to RAID 10. So I can purchase (2) more 1TB drives, break the RAID 10 and configure RAID 6. So if I have (6) 1TB drives, I should have 5TB of storage available since 1 drive is considered a hot spare? or do i have this wrong?
0
datatechdc
Asked:
datatechdc
  • 3
1 Solution
 
DavidCommented:
No a 6-disk RAID6 uses two drives for parity, so usable will be 4 disks x 1 TB.
0
 
AustinComputerLabsCommented:
Dlethe is correct:

If you use RAID 5 with a hot spare you have lost 2 disks worth of usable space - one for the the hot spare doing nothing and one for the parity space needed to recover from a drive failure.

With RAID 6 you would essentially have what would have been your hot spare being a working hot spare that will allow you to lose 2 drives at once without data loss.

So RAID 5 with a hot spare has the same overhead (loss of usable space) as RAID 6. This becomes more important as drives get larger and rebuild time with a hot spare takes longer. If a another drive fails during the rebuild on RAID 5 data is lost.

RAID 6 also has a write speed penalty because writing the two sets of parity takes time.

RAID 10 does not have the same level of write speed penalty and you can also lose 2 drives but not just any two you have to be lucky enough to lose two that are from different sets.

I hope that helps.
0
 
DavidCommented:
One big difference on R5 + spare vs. R6 is that RAID6 has a much higher tolerance and protection against data loss.

Real world, you have at least one order of magnitude higher probability of losing a block of data vs. the entire disk.   So if you have RAID5 but just one bad block on any of the surviving disks, then you have partial data loss.  (Of course that bad block will be in the center of a 500GB exchange file, so the entire file may be lost).

But RAID6 has twice the parity, so it provides protection against a bad block.  Lose one disk + have one bad block on a surviving drive, then you are protected in a RAID6 config.  You are NOT protected in RAID5.

As far as performance overhead.  get a decent controller and you may not see any performance difference except in certain benchmarks which rarely reflect real-world I/O loads.
0
 
datatechdcAuthor Commented:
i am going to be creating a dynamic disk and using it as a SQL server 2008 r2 repository. does that change anything?

do SCSI drives only come in 1TB sizes? do all of the drives in RAID 6 have to be the same size?
0
 
DavidCommented:
In general, the drives  can be different sizes, but they will end up being reconfigured to the smallest common capacity. I.e if you have all 2TB drives, and 1 40GB disk, then each disk will only be able to use 40GB.

If you are using linux or solaris zfs, then you can do all sorts of creative things to use that free capacity.   All sorts of disks out there with various sizes.  I've got some 3TB Seagate Serial SCSI drives in the lab, so obviously 1TB isn't the upper limit.
0

Featured Post

Free Tool: Site Down Detector

Helpful to verify reports of your own downtime, or to double check a downed website you are trying to access.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now