Construction of infrastructure architecture of virtualization data center in emergency management command center in certain city
Client background
The current data center structure of users refers to the traditional data center architecture, as shown in the Figure. One set of business system shall be established quickly to meet the requirements of government in rules; In the light of objective of saving investment, easily managing and saving energy and reducing emission, the VMware is selected as virtualization platform and the infrastructure architecture is required to own high flexibility to lay a solid foundation for "cloud journey".
Scheme design
Aiming at the needs of users, we select the Vblock 0 product of VCE (VMware, Cisco, EMC union) architecture set up the infrastructure platform for users. The vSphere of VMware Is used as the virtualization platform; As the hardware platform and a packaged solution through the pre-configuration and compatibility certification, the uniform computing product UCS and the uniform storage product NS1M of Cuco are provided for the users who could conduct centralized monitoring to the system on one uniform management interface and could rapidly provide computing and storage resources to rapidly pass the deployment and application of virtual machine, which completely meets the demand for rapid deployment of clients and is fully recognized by the customers.
User benefits
l Greatly shorten the construction cycle of project. The physical resource utilization increases by 60%, and the purchase expenses of hardware is reduced.
l Save the human resource used for maintenance. Faster and more flexible application deployment greatly improves the efficiency of "emergency command".
Certain scientific research institution utilizes the virtualization disaster tolerance to construct the Active-Active data center.
Client background
The user data center has used the server virtualization technology of VMware whose advantage lets users plan to transfer more production application to the virtualization environment. Users plan to set up a set of disaster tolerance data center in another machine room of the same city due to most of quite important data in scientific research institute. The most basic requirement is to achieve the automation of most actions upon the disaster tolerance switching to avoid the man-made problems; The second requirement is that the switching time shall be as short as possible; The final requirement is to conveniently conduct the test drilling of disaster tolerance.
Scheme design
The users use the storage equipment of EMC at main preparation site, and the Site Recovery Manager (SRM) of VMware could be utilized to achieve the virtualization disaster tolerance among main preparation sites by combining with the MirrorView storage replication technology of EMC.
SRM has the following advantages:
lAutomation for switching of disaster tolerance
Conduct the full-automatic operation to the virtualization platform and back-end storage on the single management interface of SRM by combining the virtualization disaster tolerance interface provided by the storage manufacturer. It is necessary to click the "confirmation" to achieve the full-automatic disaster tolerance switching.
l Save the resource investment of disaster tolerance site
One of the features for virtualization is the flexible usage of resource, so the construction scale of disaster tolerance site could be lower than that of production site. In the meantime, under the normal status, the disaster tolerance site could also apply the inquiry, test, verification and other applications of users to avoid the disaster tolerance site at Standby status all the time and to achieve the Active-Active data center architecture.
l Conduct the online disaster tolerance switching drilling whenever and wherever possible
Under the premise of not influencing the production environment, the simulation environment is completely unanimous with the production environment in the switching progress of disaster simulation at disaster tolerance station. The users could conduct the switching training whenever and wherever possible to verify the applicability and integrity of disaster tolerance environment data.
User benefits
l Extend the disaster recovery protection.
l Now, the least operation and cost are added to protect all workloads in the virtual machine.
l Reduce the recovery time
l Once the disaster happened as confirmed, press the next button to start up the recovery process for thousands of virtual machine.
l Improve the reliability of the recovery
l The replication function system standards ensures the virtual machine owns all data needed in the startup process.
lThe hardware independence avoids the fault incurred by different hardware.
l Based on the actual fault switching process, the drilling process has simple operation and could conduct frequent and highly realistic disaster tolerance drilling.
Certain financial user conducts the virtual environment backup with the elimination technology of repeating data.
Client background
A user uses 3 sets of VMware ESX service and operates over 30 applications in the production environment. Such user owns perfect data backup strategy and conducts the image-level backup to the virtual machine with the traditional backup software initially through the VMware VCB backup interface. There is a huge data volume in the backup as founded in several backup circles due to the characteristics of VMware. VMware packages the virtual hardware, OS and application software of virtual machine into a VMDK file. Except that all VMDK documents have backup in each full backup, the data volume of incremental backup is almost the same as the full backup, because the subtle operation or amendment of VM would change the VMDK document, and directly cause the incremental backup still needs the complete backup of the whole VMDK document. In the meantime, most of VM operation systems are the same and many same data are backed up, which wastes lots of backup space and occupies lots of network resources.
Scheme design
The virtualization characteristics decide the traditional backup software has no place. Backup method based on the elimination of repeating data starts dominating the virtualization backup field. We select Avamar for the backup of VM against the insufficient environment bandwidth of user. Avamar conducts the de-duplication operation to the backup data at the server terminal based on the source end and the global elimination technology of repeating data, and only a small amount of data needing backup after de-duplication reaches the backup medium through network, which greatly reduces the occupation of network resource. In the meantime, most of data in VMDK document wouldn't change. And each VMDK contains similar operation system and lots of repeating data. Avamar would conduct the global de-duplication operation when the data with backup at many source sides are deposited to backup medium in order to guarantee the backup data in the whole environment is unique. The characteristics of Avarmr decides there isn't the concept of incremental backup due to full backup upon each backup. And the user is more convenient in recovery, and the designated time point could be directly recovered upon one operation.
User benefits
l Reduce the network bandwidth needed in the backup every day, which improves the overall utilization rate of the network resources.
lThe maximum elimination rate of repeating data reaches 500:1 through the de-duplication operation.
l Reduce the mobile data volume and shorten the daily backup window.
l The full backup and the incremental backup aren't needed to be recovered successively on the basis of single step recovery.
l Reduce the backup pressure of infrastructure and prepare the larger-scale virtualization integration.
l Daily integrity inspection ensures there isn't single-point fault in the RAIN system structure and duplication.
