Typical application scenarios

Scenario: Server virtualization

An enterprise hopes to achieve the following goals by integrating the server resources of its data center through server virtualization:

• Improve the hardware resource usage of the data center.

• Control the IT resource demands.

• Reduce the purchase and maintenance costs for customers.

• Save the equipment room space as well as equipment room power supply and cooling costs.

Analysis

The enterprise requires to perform centralized management on the server resources of its data center, deploy services and application systems based on the virtualization environment, and ensure that all its branches, offices, and partners can access the service and application systems as if they were in a physical network. To achieve these goals, configure CAS CVM and use its server management and virtualization features to perform centralized management on the hardware resources and deploy the service application systems based on the virtualization environment.

Service planning

Network planning

In the CAS network, plan the management network, service network, storage network, backup network, and migration network separately. You must plan the management network, service network, and storage network at least.

Storage planning

• Local disk planning

If a server has two local disks, configure RAID 1 for the disks. If a server has more than two local disks, configure RAID 5 for the disks.

• External storage planning

CAS supports IP SAN and FC SAN storage devices. As a best practice, configure RAID 5 or RAID 6 for the storage devices.

CAS supports shared file systems that are not larger than 64 TB. As a best practice, assign 10 to 64 TB to a storage volume.

Host pool, cluster, host, and shared file system planning

A host pool can have a maximum of four HA clusters, and each HA cluster can have a maximum of 32 CVK hosts.

A cluster can have a maximum of 10 shared file systems.

CAS supports shared file systems that are not larger than 64 TB.

A shared file system can be used by one cluster.

Configuration workflow