This practice describes the procedure to create a RAID array using an EVS.

Overview of Practices

Redundant Array of Independent Disks (RAID) is a storage technology that combines multiple independent disk units to improve data storage performance and data reliability.

The RAID technology stores data on multiple disks for redundant backup, parallel read and write, and fault tolerance. Disk arrays combine to provide high performance, high reliability, and high availability.

The RAID technology works based on different RAID levels, each with its own unique data distribution and redundancy policies. Common RAID levels include RAID 0, RAID 1, RAID 5, RAID 6, and RAID 10. Each level has its own unique advantages and characteristics. You can select a RAID level according to your needs. The following table shows you the detailed information of RAID levels:

RAID Level	Level Description	The Minimum Number of Disks Required to Create a RAID Array
RAID 0	Data is evenly distributed across multiple disks to improve performance and data throughput. RAID 0 has no redundancy, and if one disk fails, the data in the entire array is lost.	Two disks
RAID 1	Data is replicated on multiple disks to provide redundancy and fault tolerance. All data is completely copied to each disk, and if one disk fails, the data is still available. However, it is less capacity efficient because twice as much storage space is used.	Two disks
RAID 5	Data and verification information are distributed across multiple disks to provide improved performance and redundancy. Data and verification information are stored in an interlaced manner, so that even if one disk fails, data can be recovered by calculating the verification information.	Three disks
RAID 6	Data and two pieces of verification information are distributed across multiple disks, and data can be recovered even if two disks fail at the same time. RAID 6 is useful for scenarios that require higher fault tolerance and data protection.	Four disks
RAID 10 (RAID 1+0)	RAID 10 is a combination of RAID 1 and RAID 0. Multiple disks are divided into two groups. In each group, RAID 1 is used to mirror data and RAID 0 is used to provide performance gains.	Four disks

Preparations Before Operation

This section describes the preparations you must make before you begin operations, including resource planning and resource creation.

Resource Planning

This practice uses a cloud server running on CentOS 7.6 64-bit operating system as an example to describe how to create a RAID10 array with four EVS disks. You must plan the resources in advance as follows:

Resource	Resource Description	Quantity	Link for Details
ECS	The operating system is "CentOS 7.6 64-bit"	1	ECS Billing
Elastic IP (EIP)	ECS must be bound to an elastic IP to download the installation package	1	EIP Billing
EVS	Data disk: 10 GB	4	EVS Billing

Creating Resources

· Create an elastic cloud server with the operating system image being CentOS 7.6 64 bit. For more information, see Creating ECS.

· Create an elastic IP and bind it to the elastic cloud server. For more information, see Purchasing an EIP and Binding ECS Instance

· Purchase four 10 GB data disks and mount them to the elastic cloud server. For more information, see Creating EVS and Mounting EVS

Operation Steps

The following shows the procedure to create a RAID array using EVS disks in the following two steps:

• Step 1: Create a RAID array using the mdadm tool;

• Step 2: Configure the RAID array to start automatically upon startup.

Step 1: Create a RAID array using the mdadm tool

In the following steps, the operating system is CentOS 7.6 64-bit. The returned information is for reference only.

Log in to the console.
Click in the upper left corner of the console and select a region. In this topic, East China 1 is selected.
Select Compute and click ECS to go to the ECS page.
Locate the row that contains the ECS you want to operate, and in the Operation column, click Remote Login .
Log in to the cloud server as the root user, as shown in the following figure:
Run the fdisk -l | grep /dev/vd | grep -v vda command to check the EVS in the cloud server and record the device name. The information as shown in the following figure is returned:

As shown in the figure, four EVS disks are mounted to the cloud server. The names are /dev/vdb, /dev/vdc, /dev/vdd, and /dev/vde respectively, each with a capacity of 10 GB.
Run the yum install mdadm –y command to install the mdadm tool. The information as shown in the following figure is returned:

If "complete" is displayed, the mdadm tool is successfully installed.
Run the mdadm -Cv /dev/md0 -a yes -n 4 -l 10 /dev/vdb /dev/vdc /dev/vdd /dev/vde command to create a RAID array with the mdadm tool.
The format of this command is mdadm -Cv RAID array device name -a yes -n number of disks -l RAID level disk 1 device name disk 2 device name disk 3 device name disk 4 device name.
The parameters are described in details as follows:
RAID array device name: The name can be customized. In this example, /dev/md0 is used.
Number of disks: Specifies the number of EVS disks in a RAID array. In this example, 4 disk are used.
RAID level: In this example, RAID10 is used.
Disk device name: Use all EVS disks you enter here to create a RAID array. Device names are separated with spaces.
The information as shown in the following figure is returned:
Run the mkfs.ext4 /dev/md0 command to format the newly created RAID array. Wherein, ext4 is the format of the file system and /dev/md0 is the name of the newly created RAID array. The information as shown in the following figure is returned:

When "done" is displayed in the returned information, the newly created RAID array is successfully formatted.
Run the mkdir /RAID10 command to create a mount path under the root directory. The following figure shows the operations and the information as shown in the following figure is returned:
Run the mount /dev/md0 /RAID10 command to mount the RAID array to the /RAID10 directory, as shown in the following figure:
Run the df -h command to check the mount result of the RAID array. The information as shown in the following figure is returned:
You can set automatic RAID array mounting when the cloud server system starts.
Run the vi /etc/fstab command to open the /etc/fstab configuration file and type "i" to enter the editing mode. Use the arrow key to move your cursor to the end of the file, and press Enter . Add the following configuration to the last line of the file: /dev/md0 /RAID10 ext4 defaults 0 0

Press Esc to exit the editing mode. In command mode, type ":wq!" and press Enter to save the settings and exit the vi editor.
Run the mdadm -D /dev/md0 command to check the information of the RAID array.The following figure shows the operations and the information as shown in the following figure is returned:

Step 2: Configure the RAID array to start automatically upon startup

Log in to the cloud server as the root user, as shown in the following figure:
Run the mdadm -D --scan >/etc/mdadm.conf command to save array information to the configuration file and run the vi /etc/mdadm.conf command to open the file to check its content. The information as shown in the following figure is returned:
In the mdadm.conf file, enter the detailed information of the new RAID array. Move the cursor to the header of the information saved in the previous step, and press Enter . Then, use the arrow key to move up one line and add the following configuration to the empty line:
DEVICE /dev/vdb /dev/vdc /dev/vdd /dev/vde
Press Esc and type ":wq!". Press Enter to save the settings and exit the vi editor.
Run the more /etc/mdadm.conf command to check whether the mdadm.conf file is successfully modified.

If you can see the information added in the previous step, the modification succeeds.

Elastic Volume Service

Building RAID Arrays using EVS