What is RAID 0, 1, 2, 3, 4, 5, 6 and 10 (1+0)?

hi guys this is Duncan from dong spoke comm and it's time for a tutorial today we're going to talk about raid redundant array of inexpensive disks or redundant array of independent disks depending on who you speak to and raid allows you to have a group of physical disks provide itself as one logical disk on your system with having magic going on in the background so you can have increased speed you can have fault tolerance you can ever done and see all that sort of stuff all going on in the background but you're only seeing one area of usable space in your operating system there's several different configurations and levels and there's hardware and software based versions not going to talk about those I'm going to talk about the raid levels that you can have today and I have raid 0 1 2 3 4 5 6 and 10 now 10 is a nested raid level or a hybrid raid level there's also raid 150 51 53 unloaded them but I'm only going to go through the basic ones 0 to 6 and 10 even then some of those are pretty obsolete now but worth knowing anyway and the nested ones are combinations anyways so without further ado let's get started with raid 0 raid 0 uses a form of striping and there's three different versions that used really in raid there's bit striping bite striping and block striping what it generally means is splitting data up across multiple areas so as you can see here if you can imagine these to be physical hard disks these little cylinders here and my data is on these physical hardest imagine that they're 1 terabytes each and going on they're all going to be 1 terabytes each and my date is on here now a 1 a 2 a 3 a 4 a 5 a 6 measuring that's a table in Excel or something like that is spread across this disk why is that done like that well the advantage of that if I was to say for a person to write down A to Z on a piece of paper it's going to take them some time but then ask to people to write a to Z but one of them to write a to M and the other person's write m2z the then that's going to take a lot quicker time than it is for one person to write eight is it and that's the advantage of having two drives you get much faster write performance but if I was to kill one of those people so if I killed the guy writing a to M I've lost the alphabet and I need the alphabet so if you were to take out one of these discs or one of these discs died you've lost all of your data and because it was treated as one logical drive you're not going to access your data it's gone that is pretty much raid zero so not ideal for a mission-critical system but provides speed increase so this one is raid one and if you can imagine my data is on the left hand side a one to a fall on this terabyte hard disk it's quite self duplicated over to disk one a one's a four so what this means is that a I will not have as much space as I'd like with my two terabytes in total I would only have one terabyte of usable space but I will have the reliability of if disc 1 for example was to fail or die it could or the RAID controller could simply move over to the working version I could then put a new hard drive in and it would then start duplicating itself again onto the new drum so this is ideal for reliability performance won't be affected because it's effectively only going to one of the drives well maybe with a hardware RAID controller some of them have a cache that you might be able to use which may have a small performance boost but this is mainly designed for reliability so now on to raid 2 and this is a little bit more complex and it does require a minimum of three drives so here we have four terabytes of usable storage so four different drives are my records are put across all of them my data is striped at a bit level so that means that the drives need to work in a certain way I believe I'm not going to go into the technicalities but they they have to work in a certain way for them to all work together in the striping and we also have three dedicated drives for error correction so what that means is that when data is written it will calculate the error correction code also known as Hamming odhh or ECC and it will then write it to these discs but when it comes to breeding from the disk it will then need to read the corresponding code and then if any adjustments need to be made it will do it on the fly so you may get that sequential read and write performance benefit by having these all working together random will not have a benefit but you do get the error correction but with raid 2 it's not really available anymore nobody uses it error correction is in hard drives nowadays anyway so it's a very old obsolete solution but it's good to know how it works so we have that 4 terabytes of usable space and 3 here in this case dedicated to error correction so Row 3 involves storage using striping with bytes which similarly to raid to require some synchronous between the disks these will have to spin in sync for it to store and it also contains a parity disk loose parity bits are stored in here and there rules that can define information on the records that are stored on the other three so you'll have 3 terabytes in this case of usable storage and then the extra parity disk which can contain rules so if disk 1 was to die it would be able to reconstruct that disk based on the records it has and the rules it has I won't go too technically into parity but it works that way these are minimum of three drives to operate and you'll get good sequential read and write because these are working synchronously but random read and write will have worse performance so this might be good for large video editing but again this is not a commonly used method it's just something to let you know how it works so so dedicated parity disk three data disks stored on a byte level synchronous here we are we've raid 4 which kind of performs similarly to make to raid 3 raid 4 is actually stored or striped at a block level and has that dedicated parity disk so you're looking at good random reads because the data blocks destroyed and bad random writes because for every write has to write to the parity disk so you are getting that slight delay when you're doing that it's far more efficient to go with what we're going to talk about next which is raid 5 this is something that's not very commonly used but it doesn't require that spinning in sync but yeah let's go ahead and talk about raid 5 so here we are with raid 5 and this time instead of having a dedicated parity drive which would make you lose a terabyte of space over here we have distributed parity which again will allow you to lose a terabyte of space but you can actually have more space efficiency here by storing across all of the stuff so again disk 1 was to die you can replace that disk and it can get reconstructed by the missing records that are available on the other drives and through the rules of that parity disk so this is probably the most cost-effective solution you do lose that space but you get good performance too because it is doing block-level striping and yeah row operations might be a little bit slow because again it does have to write that parity information but it's pretty good for things like databases and here we are with raid 6 this gets a little bit more complex this is actually dual parity so you're having essentially to parity dedicated discs distributed across all of your disks now this is probably more useful if you have a lot more disks and as shown here because two disks can actually fail AB your entire setup you can replace those and again using those parity rules and all the records that are available you can then get them replaced and it will be efficient so again can handle two disk failure where you can configuration is complex though and yeah that's used through the block striping system too and here we are with raid 10 which is also known as a nested level one plus zero the reason for this is because we have a raid 1 setup in which are talked about before where you are mirroring data but is then striped across so in this case for this disk here to die your system won't screw up because you have that backup with the first drive-in raid one and it's also giving you the level of performance with the raid zero stripe but it's only giving you half the usable space in this case so this four terabytes here would only give me two terabytes but it's a very efficient solution because if you're adding more drives an entire pair of drives would need to fail in order for this drive to mess up because there's a nested level and then you have that entire raid zero giving you that added performance that you'd require so this is a highly regarded system if you can afford it and can afford the space loss then it's worth having a shot with rate 10 so that pretty much sums up all the different raid levels the most common and some obsolete if you like this video feel free to give a thumbs up if you have any comments about any of the things I've talked about today feel free to let me know if I said anything wrong feel free to correct me and I will see you guys in the next video

Channel: mobilephone2003

Views: 992,799

Rating: 4.9121442 out of 5

Keywords: hdd, hard, disks, levels, level, Hard Disk Drive (Invention), computer, speed, redundancy

Id: wTcxRObq738

Length: 9min 57sec (597 seconds)

Published: Mon Sep 09 2013