their Filer did not have the capability to hot swap the bad disk with another good disk. Instead, RAID 5 rebuilds the disk when it fails, which uses up a lot of CPU and time. With RAID 0+1, which their SAN device uses, the drive that failed can be pulled out and replaced. Not so in RAID 5.
Your friend needs to do some studying before being a part of such a decision. NTAP absolutley does hot swap a bad drive. I does rebuild the contents of the failed drive to the new drive (which was sitting ready for use) using RAID parity techniques. Sys admin controls the amount of CPU resource allowed to be used in the rebuild. It can be as low as 10% or as high as 90%.
NTAP does not use RAID 5. It uses RAID 4, which also allows the user to grow a RAID array simply by adding a drive with no file rebuild as required by RAID 5.
"RAID 0+1" is not RAID at all. Though I am not sure what "0+1" means, as that is not a RAID term, it sounds like they are using Mirroring, which is very expensive and also slow, as all writes are duplicated.
What your friend lost- Speed, Reliability, Simple sys admin, multi-protocol (NT/UNIX) securit, SNAPSHOT/RECOVER, addition of clustered configurations, simple growth.
What your friend gained- Nothing, but probably at a higher price.
This is more than just my opinion. |