Distinguish between basic disk and dynamic disk RAID disk redundant array

One 、 What do you mean RAID

RAID Our English full name is Redundant Array of IndependentDisks

The Chinese name is

Redundant array of independent disks

RAID The original intention of the technology is to provide high-end storage functions and redundant data security for large servers .

In the system ,RAID It is regarded as a logical partition , But it is composed of multiple hard disks （ At least two ）. It works by simultaneously

Store and read

Data to greatly improve the data of the storage system

throughput

（Throughput）, And in many RAID There are more complete mutual checks in the pattern / Measures for recovery , Even direct mirror backup of each other , So as to greatly improve RAID The fault tolerance of the system ,

It improves the stability and redundancy of the system

, This is also Redundant The origin of the word .

Two 、 Common disk redundancy levels

• RAID0

• RAID1

• RAID5

• RAID10

• RAID01

• RAID50

3、 ... and 、 meaning

RAID0

• RAID0 With
  Zone
  Form stores data on two or more physical disks , Data is alternated 、 Allocate these disks equally . Read and write in parallel , At all levels ,RAID 0  The speed of
  The fastest
  Of , but
  No redundancy
  .
  

• 10110001
  

RAID1

• RAID 1 Yes, it will
  Same data
  Copy to two groups of physical disks at the same time . If one of them fails , The system can continue to use the undamaged disk ,
  Highest reliability
  , But its disk utilization is only 50%, It's all RAID Disk utilization in level
  The minimum
  A level of .
  

• 10110001
  

RAID5

• RAID5 It is to write data to the disks in the array , Put the
  Parity data
  Stored interactively on each hard disk . Any hard disk is damaged , It can reconstruct the damaged data according to the check bits on other hard disks .RAID 5 In an array
  At least three physical drives are required
  , The utilization rate of hard disk is n-1/n. Best value for money .
  

• Parity check (Parity Check) It's a kind of
  Verify the correctness of code transmission
  Methods . According to the bits of a set of binary code transmitted “1” The number of is odd or even to verify . Using odd numbers is called odd check , conversely , It's called even parity . What kind of calibration should be used is prescribed in advance . Usually, a special parity bit is set , Use it to make this set of code “1” The number of is odd or even . If odd check is used , When the receiver receives this set of codes , check “1” Whether the number of is odd , So as to determine the correctness of the transmission code .
  

• RAID5 Parity in
  

RAID10

• RAID 10 yes
  RAID 1
  and
  RAID
   0 The combination of , In all RAID In the hierarchy ,
  performance 、 The protection function and capacity are the best
  Of . Most of the time ,RAID 10 It can withstand the failure of multiple disks , So the system
  Higher availability
  , But in combination RAID 1 and RAID 0 Advantages at the same time ,RAID 10 There are also and RAID 1 The same redundancy , Disk utilization is too low .

• It applies to
  High security requirements
  、
  High load application scenarios
  , The default configuration of storage system high-end applications generally adopts RAID 10 Pattern .

• 

RAID01

• RAID01 And RAID10 be similar , All are RAID0 and RAID1 The combination of . The difference is RAID01 Yes, first RAID0, After the RAID1. Both provide RAID1 Same data security , Also provided RAID0 Approximate storage performance
  

• RAID10 And RAID01
  

RAID50

• RAID 50 yes RAID 5 and RAID 0 The combination of , Inherited RAID 5 High disk utilization and RAID 0 Advantages of high speed ,RAID 50 Higher fault tolerance , Because it allows one disk in a group to fail , Without data loss . And because the parity is in RAID5 On the subdisk group , So the reconstruction speed has been greatly improved .RAID 50 Its advantages are higher fault tolerance and faster data read and write rates .

• 

Four 、 Common volume groups

Simple volume

• The simple volume is in
  Separate dynamic disks
  A volume in , It is similar to the partition of the basic disk . But it is
  There is no space limit
  as well as
  The limit of quantity
  . When the space of simple volume is insufficient , You can also expand the space by expanding the volume , And this will not affect the data in the slightest .

Cross region volume

• A cross region volume is
  One contains multiple disks
  Volume of space on （ most 32 block ）, The order in which data information is stored in a spanned volume is to fill the first disk and then gradually store it in the following disk . By creating a spanned volume , We can
  Allocate the free space in multiple physical disks into the same volume
  , Utilized resources . however , Cross region merging
  Cannot improve performance or fault tolerance
  .

Band volume （RAID0）

• The tape roll is made of
  Two or more hard disks
  The composition of , Is that the empty size contributed by each hard disk must be the same , It is a dynamic volume , Must be created on a dynamic disk .
  Use of striped roll  RAID-0
  , So that we can be in
  Distribute data on multiple disks
  . Striped volumes cannot be expanded or mirrored , And does not provide fault tolerance . If it contains one of the Striped volumes
  A disk has failed
  , be
  The entire volume does not work
  .

Mirror volume （RAID1）

• The mirrored volume is a volume with
  A simple volume of identical copies
  , It needs two disks , A piece of data in operation , A copy that stores exactly the same , When
  When a disk fails
  ,
  Another disk can be used immediately
  , Avoid data loss . Mirrored volumes provide fault tolerance , But it does not provide performance optimization .

RAID-5 volume

• So-called RAID5 The volume contains
  Striped volume of parity value
  ,Windows Server 2016 Add a parity value for each disk in the volume set , In this way, the superior performance of the strip coil is ensured and , It also provides fault tolerance .RAID5 volume
  Include at least 3 Block disk
  , most 32 block , In the array
  Any disk failed
  when , Fine
  Calculate from the information in the other two disks
  , And recover the data in the failed disk .