Hi, I'm Calvin Mangubat, I'm a technical trainer with Paloalto networks. For this session we'll be talking about the Redundant Array of Independent Disks or RAID. We'll talk about how it works, what it does to your data. So RAID, is a storage technology that you'll often find on servers and local actors. It's used whenever you have the following requirements; speed of access to data and fault tolerance. You care about your data enough that you need to get at it fast, and you want to make sure that you never lose that data. Now RAID, let's just highlight that. Speed of access and fault-tolerance. So, if you lose one data will still be there. Now the R in RAID stems from the fact that your data is now stored across multiple disks, okay. So depending upon the RAID level you can have two, you can have three, you can even have more, okay. So there are different implementations of RAID and these are known as RAID levels. Now the techniques that are applied to your data are typically one or two or a combination of the following. The first is known as striping. So with striping what happens there is your data is broken up, and then it is spread across your discs. So they're written in parallel on both disks so that's how you achieve your speed. However it's not very good for fault tolerance because if you lose one disk then your data is lost. So there is another technique known as mirroring. Now, as the name implies it's simply a matter of writing the same data in two locations. So if you lose one disk, your data will still be available in the other. With mirroring read speed is good, writing however isn't so great because you're now writing in both locations. So you don't have the same advantage that you would have we just plain striping. So, with striping this is the default that you would find in what's known as RAID zero. Mirroring, this is what's used in what's known as RAID level one. Now there are other RAID levels that actually use a combination of both techniques and you also have some that leverage one of these other techniques and it basically improves upon it. Let's take for example what's known as RAID five. In RAID level five it's based on striping but you're now adding the concept of parity. So with RAID five you'll need a minimum of three disks. So you'll have for example data here, data here and you have the concept of parity information. Now parity information allows you to rebuild the data if data is lost. Now parity is calculated using a logical operator known as XOR, okay. Now, let me give you a very basic example of how this operation works and how it can actually be applied to your data. So with XOR if you take two bits of information, and remember when data is stored in a disk it's stored as ones and zeros, right? So if you compare two bits of information, you run a sore operation to it, if they're the same the parity value will be zero. If they are different then the value will be one. So if for whatever reason this information disappears, all you have is this the fact that the parity value zero you already know the data that is missing is zero. Let's expand on this a little bit with the little more complicated bit of information. Let's say for example 011, 110 like so. So, if you were to calculate parity for this value, so zero and one that's different, one and one is the same, one and zero is again different, so it's 101. For whatever reason if this data disappears, you can still figure out what this was originally by doing a comparison. So one, zero that's one. One and zero that's also a value of one. Since they're the same, remember if they're the same it's a zero. It becomes a zero. So as you can see here the data comes back. That is how you bring data back using your parity information. So let's summarize. So RAID is a storage technology that allows you to speed up read and write access to your data, it also provides fault tolerance. Techniques that are applied to your data are striping, which allows you to spread data across multiple disks. Mirroring, which allows you to write your data in multiple locations, identical data not information, and if you use striping, if you use parity information you can use your striped information to bring back data from disks that have failed or have been removed. Now with other RAID levels you will see a combination of striping, mirroring, and parity, and they will be used in combination. Thank you very much for your time and I hope you will watch our future light-board sessions.
