Microsoft cloud Computing Design and practice Virtual data Center

A considerable number of enterprises in the information project lack of planning awareness often will pursue a lot of growth in equipment, this time high input often can not make all resources can be used efficiently, but because of this cloud computing server virtualization Technology is the solution to this dilemma is the direct means and methods.

Example: Mr. Huang is the head of operation and maintenance of the Informatization Department of a state-owned holding enterprise. Has participated in the Beijing high-end office building Network and Data center planning and implementation. During the current business, the company is responsible for building and rebuilding existing data centers. He proposed that the use of server virtualization technology to improve the efficiency of the existing equipment, through the virtualization technology to achieve a unified development environment management to reduce the complexity of operation and maintenance.

List of existing and IT system equipment:

2U server Dozens of units, 15T storage server 2;

The development test environment occupies 7 servers, serving 4 IT projects that are being developed respectively;

The Production Environment Web application System 4 sets, the business System 2 sets occupies 8 servers and 1 storage servers;

The remaining devices are other small projects, backed up or hot standby.

Environment after virtualization:

Host 1 Servers mount 1 storage as master, provide virtual machine 6 virtual machines for development test environment, and reserve backup of storage server.

Implementation steps:

Combined with Mr. Huang's real-life experience, we can put the implementation of virtualized data centers in the same way as other IT projects to go through the start-up, plan, execute, monitor, and close the five processes, but to ensure that our virtualized data centers are available for a long time and the environment is stable, a complete set of design and planning processes are in place, as shown in Figure I.

Phase one: Start the Virtualization implementation project

Plan the virtualization scope, according to the company's business and technical requirements to plan the physical location of the server, in this case, because the small data center has been centralized storage so you can ignore this problem. Define virtualization goals and implementation schedules and related principals, and comb through the virtualization process or service list. Completes the equipment and the software application the reorganization and the record, provides the cushion for the quantification demand.

Phase II: Planning requirements

Analysis of relevant statistics, you can use third-party tools to calculate performance requirements, such as the use of PAL tool analysis of Windows system performance records. The requirements of the application can be generated by the actual operational records of the original application, and the specific requirements are planned according to different procedures. A particularly important point here is to choose a backup method.

The first thing to distinguish is the backup content, whether for individual application backups, virtual machine backups, or backup physical hosts. Especially in this case, the backup requirements in the test environment are mainly for the backup of the virtual machine, so the storage server that is mounted must reserve some space.

In addition, for performance, downtime, storage space, ease of recovery and operational complexity, consider the data loss program and backup interval. For example: implemented through the Windows Server Backup tool, the Hyper-V management console, and so on.

Phase III: Design options

1. Select the configuration of the physical host to determine whether a new hardware device needs to be purchased, based on the full use of existing hardware devices, taking into account cost, performance, availability, stability, and manageability. such as the Map tool for evaluation. Similarly, the host is placed in the IDC room, self-built computer room is still to be considered, in this case because it has been used since the building of the room so that the temporary can not be considered.

2. Virtual machine corresponding physical host

Need to understand the physical host resource utilization target, hardware capacity. Avoid placing virtual machines that are consuming system resources on the same physical host. Use tools include: SCVMM 2008 R2 Intelligent Placement, etc.

3: Design High Availability

Selecting a virtual machine for failover clustering is the establishment of a failover cluster between Hyper-V hosts or between hosts and virtual machines. Tools and methods: have Hyper-V host failover cluster, Live migration work mechanism, network redundancy.

4: Designing Your Storage infrastructure

Depending on your needs, choose directly attached storage (DAS), NAS, San, and so on, whether the virtual machine uses dynamic or fixed-size disk space and needs to be planned according to the environment. After all, the development test environment in this case is not very urgent for storage requirements.

5. Design Network Infrastructure

What's clear here is that Hyper-V provides three types of virtual machine choices, DMZ, Intranet, Internet, or VLAN functionality.

Phase IV: Confirm that all methods are feasible

Verify that the identification in the above steps conforms to the enterprise requirements and confirms the application requirements and physical host infrastructure design.

Phase Fifth: implementation and monitoring

After detailed design and planning, implementation of Virtual data Center Server Monitoring management, implementation of virtualization preparation has been completed, we can through the host virtual machine management software system Center Virtual Machine Manager Management monitoring virtual machine performance, and targeted to the virtual environment monitoring and adjustment.

Summary of the case in this case, has a distinct feature: a large number of enterprises in the information technology projects in the lack of planning awareness will often pursue a lot of growth in equipment, this time high input often can not make all resources can be effectively used, But because of this cloud-computing server virtualization technology is the direct means to solve this dilemma, it can more exert its control and improve the utilization of equipment, improve the enterprise IT department ROI. However, it is necessary to realize that virtualization is not an inexplicable cost, but a serious scientific planning and design to achieve the desired goal.