Question : Problem: Fault tolerance

This is a question from a MCSE book


You administer a server called Server1 which has a single physical disk that is configured as a simple volume

You plan to store the files for a large database on the server. You plan to install additional physical disks on the server

You need to configure the disks on the server so it provides fault tolerance for the operating system and the database files.

The answer says install 3 additional disks and create a new RAID-5 volume, and place the database files on the new volume

And

Install another disk and configure the simple volume as a mirrored volume


My question is: - Why three extra disks for the RAID-5,

And

Why do u need the Additional disk for the mirrored volume

Answer : Problem: Fault tolerance

3 is the minimum number for a RAID 5 array. Reason being that data is stripped across 2 disks and parity check bits are stored on the 3rd. In Theory if a disk fails then you can recreate the data from half the stripe and the parity. If Parity fails, you still have the data. In reality, this is short sighted and you should use more disks to provide faster recovery and hot standby disk to give zero impact in case of a failure. You can also stripe the parity across all drives to provide slightly better recovery.
You have no resilience in a single drive. Therefore, the mirror pair allows for a single drive to fail without impact.

Hope this helps.... or where you asking something else?
Barny