RAID (Redundant Array of Independent Disk independent redundant disk array) technology is the University of California at Berkeley in 1987 raised, initially to a combination of small low-cost disk to replace large expensive disks, and hope does not make access to the data when disk failure developed a certain level of data protection technologies suffer. RAID is a kind, in the operating system as a separate storage device fails by a large plurality of redundant array of inexpensive disks thereof. RAID can give full play to the advantages of multiple hard drives, you can upgrade the hard drive speed, increased capacity, enough to provide fault tolerance to ensure data security, easy management advantages, can continue to work in the absence of any hard disk problems, and will not affected by damage to the hard disk.
RAID is a Redundant Array of Indepent Disks (redundant array of independent disks) acronym, is the most commonly-used four RAID RAID 0, RAID 1, RAID 5, RAID 10, explained below by way of illustration and distinction characteristics of the four RAID .
In the latter illustration, the following identity is used:
- A, B, C, D, E and F - represents a block of data
- p1, p2, p3 - parity block represents
RAID 0
RAID 0 characteristics:
- It requires a minimum of two disks
- Distribution data strip formula
- No redundancy, the best performance (no storage mirroring, parity information)
- It can not be applied to data security for high occasions
RAID 1
The following is a RAID 1 characteristics:
- It requires a minimum of two disk
- Provide data redundancy block
- Good performance
RAID 5
RAID 5 characteristics:
- A minimum of three disks
- Distribution data strip form
- For redundancy to parity
- Read more for less write scenarios, performance and data redundancy is the best compromise
RAID 10
RAID 10 (also known as RAID 1 + 0) Features:
- Minimum of four disks
- Press RAID 0 divided into two groups, the two groups were then mirrored RAID 1 by way
- Both redundant (mirrored memory provided) and performance (data stripe-shaped distribution)
- In practice more commonly
RAID 0即Data Stripping（数据分条技术）。整个逻辑盘的数据是被分条（stripped）分布在多个物理磁盘上，可以并行读/写，提供最快的速度，但没有冗余能力。要求至少两个磁盘。我们通过RAID 0可以获得更大的单个逻辑盘的容量，且通过对多个磁盘的同时读取获得更高的存取速度。RAID 0首先考虑的是磁盘的速度和容量，忽略了安全，只要其中一个磁盘出了问题，那么整个阵列的数据都会不保了。
问：RAID0至少几块盘？
答：RAID0最少要两块硬盘才能实现。
RAID 1
RAID 1，又称镜像方式，也就是数据的冗余。在整个镜像过程中，只有一半的磁盘容量是有效的（另一半磁盘容量用来存放同样的数据）。同RAID 0相比，RAID 1首先考虑的是安全性，容量减半、速度不变。
问：RAID1至少几块盘？
答：RAID1最少要两块硬盘才能实现。
RAID 0+1(RAID 10)
为了达到既高速又安全，出现了RAID 10（或者叫RAID 0+1），可以把RAID 10简单地理解成由多个磁盘组成的RAID 0阵列再进行镜像。
问：RAID0+1至少几块硬盘才能实现？
答：RAID0+1至少需要4块盘。
RAID 3和RAID 5
RAID 3和RAID 5都是校验方式。RAID 3的工作方式是用一块磁盘存放校验数据。由于任何数据的改变都要修改相应的数据校验信息，存放数据的磁盘有好几个且并行工作，而存放校验数据的磁盘只有一个，这就带来了校验数据存放时的瓶颈。RAID 5的工作方式是将各个磁盘生成的数据校验切成块，分别存放到组成阵列的各个磁盘中去，这样就缓解了校验数据存放时所产生的瓶颈问题，但是分割数据及控制存放都要付出速度上的代价。
问：RAID5需要几块硬盘?为什么损失一个盘的容量?
答：至少3块。
RAID5把数据和相对应的奇偶校验信息存储到组成RAID5的各个磁盘上，并且奇偶校验信息和相对应的数据分别存储于不同的磁盘上，其中任意N-1块磁盘上都存储完整的数据，也就是说有相当于一块磁盘容量的空间用于存储奇偶校验信息。因此当RAID5的一个磁盘发生损坏后，不会影响数据的完整性，从而保证了数据安全。当损坏的磁盘被替换后，RAID还会自动利用剩下奇偶校验信息去重建此磁盘上的数据，来保持RAID5的高可靠性。
