Storage is a very important part of
fault tolerance. If something were to happen to a
company's data such as a disk failure that results in data loss, then that could
have a serious impact on how the company performs. That's why we need to make sure that if
a disk does fail, that no data loss would occur. And one of the best ways to prevent data
loss is RAID, RAID stands for redundant array of
independent disks. In a RAID setup, the data is copied on
multiple disks. So that in the event of a disk failure, no data would be lost. Now there are four common types of RAID.
There is RAID 0, RAID 1, RAID 5, and RAID 10. Now RAID 0 is not fault tolerant. In
fact RAID 0 shouldn't even be called RAID because not only does it not
provide fault-tolerance, it actually increases the chance for data loss. Because in a RAID 0, the data is not
duplicated, but it's actually spread or striped across two separate disks. So it's just one of these disks fails, or
if you decide to destroy yourself with a hammer, then all the data will be lost. So the only reason why you would want to
use RAID 0 is speed. Because when you have 2 disk
controllers working instead of 1, then accessing data is much faster. Now RAID 1 is fault tolerant. In a RAID 1 set up the data is copied on
more than one disk. So disk 2 will have the exact same copy of the data as disk
1 So in the event of a single disk failure,
such as getting destroyed by a laser beam, then no data loss would happen because
the other disk would have a duplicate copy. Next we'll talk about RAID 5. In order to use RAID 5 you need to
have 3 or more disks. RAID 5 is probably the most common setup that is
used because it's fast and it can store a large amount of data. So in a RAID 5 setup, data is not
duplicated, but its striped or spread across multiple disks. And in addition to the data, there is
another very important piece of information that is being evenly spread
across all the disks, and this information is called parity, and parity is used to rebuild the data
in the event of a disk failure. But there is a downside to RAID 5,
because since the equivalent of an entire disk is used to store parity, it
reduces the collective amount of data that can be stored in this array. So for example if all 4 of these
disks were 1 terabyte each, that totals 4 terabytes, but in a RAID 5 set up, the total amount
that will be used for data storage would be 3 terabytes, because the equivalent
of 1 entire disk would be used to store parity. And finally there is raid 10 and RAID 10 is basically what the name
says, its combining RAID 1 and RAID 0 together, and you need to use a minimum of 4
disks. So in a RAID 10 set up, a set of 2 disks
are mirrored using a RAID 1 set up. Then both sets of the two disks are
striped using RAID 0. So RAID 10 benefits from the fault
tolerance of RAID 1 and the speed of RAID 0. But the downside in a RAID 10, is that you can
only use 50% of the capacity for data storage. So if you are using four disks
in a RAID 10 setup, you can only use two of them for actual storage.
