The term RAID is the acronym of Redundant Array of Independent Disks. This is a comprehensive guide on various levels of the RAID, like RAID 0 Raid 1, Raid 5, and Raid 10. In practice, the RAID does not indicate anything other than a technique that allows the controller to manage the different storage units and divide the data among the multiple disks present. This increases performance, security and even tolerance against any faults.
How Does RAID Work?
Thanks to the lightning-fast speed offered by the solid state units (SSD) and continuous technological evolutions, we have much superior hardware solutions available for our computer. Despite these important improvements, the need to adopt particular strategies that are able to avoid the loss of valuable information, due to faults or sudden malfunctions of the disks, has not yet been completely eliminated.
Since the management of performance, security, and tolerance against possible failures totally depends on the chosen strategy, it would be appropriate to first understand how many and what are the types of existing RAID. In practice, there are in fact the basic types of RAID, ranging from level 0 to level 7, and RAID types nested, which are nothing more than the basic RAID types combined with each other in such a way as to exploit the characteristics of one or of the other type.
For these reasons we will try to explain how the RAID works starting first from the commonly used basic RAID types, especially at home or in the office, then move on to the more advanced RAID types, exploited above all in the business environment.
RAID 0 (striping) Explained
The RAID type 0 level, also known as striping, divides the data into equal blocks, each of which is then written on a different disk. In this case, it is necessary to use at least two disks, while the actual capacity is equal to the capacity of the smaller disk multiplied by the total number of disks used.
As for the benefits of RAID 0, it is worth mentioning the ease with which this can be implemented, in addition to the performance almost proportional to the number of discs used. In this case, there is no guarantee of data recovery against any hard disk failures, in fact, if a disk is broken all data will be immediately lost. The probability that a disk can fail, increases proportionally to the same number of disks used.
RAID 0 is, therefore, a type of RAID addressed to those who must process files of different nature, ie video, audio or images, and those who need to use all those special programs that need to transfer large amounts of data. In contrast, RAID level 0 is not recommended in all those cases where data security plays the most important role.
RAID 1 (mirroring) Explained
In RAID 1, in fact sometimes it is also called mirroring, the data are first written on the primary disk and then be replicated on one or more secondary disks. In this case, it is necessary to use at least two disks while the effective capacity is equal to the capacity of the disk that has the smaller size. Contrary to what happens with level 0, in the RAID 1, it is also admitted the failure of all the disks present but due to the mirror image of one disk already exists, you are protected against any kind of hard disk failure.
The RAID level 1 is, therefore, the simplest RAID configuration as it not only guarantees data replication together with tolerance against possible failures (in this level, in fact, if you damage a disk, just replace it, even in hot swap mode, to restore lost data), but in some cases this level also allows a slight increase in reading performance, since at least in theory the controller can simultaneously make multiple readings.
As for the disadvantages, it must be said that the RAID level 1 is the one that has the worst management of available space, in fact, the total capacity will be equal only to that of the disk with the smaller size. Moreover, since data must be written simultaneously on all disks, write performance obtainable with a level 1 RAID will be comparable to those obtainable from a single disk. As a result, there will be no visible improvement in performance in this case.
Level 1 RAID is therefore recommended for all those applications where data security plays a key role.
RAID 5 Explained
Thanks to the use of the parity bit, the RAID 5 can be considered the most suitable RAID configuration for any use, as it offers both increased performance and greater data security. In this case, however, the minimum number of disks increases to three while the actual capacity is equal to that of the smaller disk multiplied by the total number of disks minus one.
For example, if you want to build a RAID level 5 with three disks of 2 terabytes each, in this case, the actual capacity will be 2 TB x (3disks – 1 disks) = 2 TB x 2 disks or 4 terabytes.
Given the ability to write and read on multiple disks at the same time, in RAID 5 the performance can be increased depending on the number of disks used, just as in RAID level 0. However, despite the high speed in reading the data together with a medium speed in writing them, if a disk fails in the RAID level 5 the general performance will be affected in a more or less serious depending on the controller used. Even though it is a fairly simple operation but the reconstruction of the entire RAID 5 system, in the event of faults, can take a very long time. However, there is a version called RAID 5 Enhanced which, thanks to the use of an additional disk, allows to reconstruct the RAID system more quickly as soon as one of the disks has been damaged.
RAID 10 (1+0) Explained
RAID 10, which can also be referred to as RAID 1 + 0, is the simplest nested RAID type. In this case, at the base of the RAID 0 system, disks are installed according to in RAID 1 configuration.
Thanks to the use of this particular structure, the RAID level 10 allows you to have very high performance, based on the number of branches in the RAID level 0, with an equally high level of security, based on the number of branches present in RAID level 1.
In this case, it is, therefore, possible to use all those applications that require high performance and, at the same time, fault tolerance. As far as the number of disks is concerned, at least four disks are needed in RAID 10. While the actual capacity is equal to that of the smaller disk multiplied by the total number of disks present then dividing all for two.
By comparing the RAID level 10 with the RAID level 0, we can therefore say that the level 10 has a tolerance against faults much better (in fact in level 10 is allowed the failure of two disks, as long as they are not part of the same Level 1 branch), while, when compared to level 5 RAID, level 10 RAID does not need to do any particular calculations to store data. Also, if you break a disk in RAID level 10, in this case, the general performance will not be affected significantly, contrary to what happens instead of RAID level 5. On balance, it could, therefore, be said that the only disadvantages of the level 10 RAID are represented only by the cost and the storage space that is halved.
How is RAID Used?
As already mentioned, RAID is linked to the use of a specific controller. The latter is more and more often integrated into motherboards and is therefore managed through the BIOS, otherwise, it can also be added through the purchase of an appropriate expansion card. Precisely for these reasons, and thanks to the possibility offered by some operating systems, it is possible to implement and manage the RAID even only via software.
However, this mode obviously does not offer the same performance that can be obtained from the mode made by hardware. In fact, the software RAID raises several limitations, such as the mode with which it can be started, the supported RAID configurations, and the ‘impossibility to perform a hot replacement of the disks, because all the operations necessary to implement a RAID system must be delegated and managed by the operating system of the moment, which in turn already uses the processor and RAM for other purposes far more traditional.
In short, the RAID offered by these controllers, integrated or dedicated, does not allow to make a real RAID system, but rather a middle ground between the software mode, which is still better than nothing, and the hardware, which in addition to a good controller, it would also need a certain amount of RAM and a processor that takes care only of making the necessary calculations, especially if you want to use the most demanding RAID configurations. Therefore, to implement a RAID in the true sense of the word, the best solution almost always consists of buying an ad hoc product: the NAS.
RAID, therefore, is a technique that offers several advantages to improve the security of storage systems, such as centralized data management combined with reliability and independence from the operating system used. However, when we want to build a RAID configuration, we must also take into account several negative aspects such as competition from the cloud services, the cost to implement it, the difficulties that are encountered in configuring it, and not less, also the noise that is emitted from all the disks when they are working.
Moreover, as efficient and effective as it may be, there is also the fact that a RAID system will never be able to prevent any damage caused by external factors, against which there will always be little to do. Once arrived at this point, we hope you now understand what RAID is, how RAID works, and how to use RAID.