Key points of construction and supervision of cloud computing data center

Cloud Computing Data Center is a data center that supports cloud service to realize user transition to customer, and it is a concentrated embodiment of a series of new technology centralized application and business service operation Management. Cloud computing data centers adopt new technologies such as virtualization, automation, parallel computing, security policies, and energy management to address problems such as increased costs and excessive energy consumption in data centers, through standardization, modularity, The architectural approach to dynamic Flex deployment and self-service services enables agile response to business services and on-demand access to services.

1, the composition of cloud computing data center

Cloud computing data centers are essentially comprised of cloud computing platforms and cloud computing services. Cloud computing services include applications, software, tools, and computing resource services that are provided to users through various means of communication, and the cloud computing platform includes a secure, reliable and efficient operating hardware and software platform to support these services. The cloud computing platform integrates the hardware and software of one or more data centers, forms a layered virtual computing resource pool, and provides computing, storage, and network communication capabilities that can be dynamically provisioned and smoothed to support the implementation of cloud computing services.

Cloud computing services are external implementations of the cloud computing hub, including services such as the Internet (Internet), application software (SaaS), System Platform (PaaS) and computing Resources (IaaS), which are characterized by no upfront investment, on-demand rental services, simple access, and safe and reliable use, Users of different sizes can dynamically expand their service content as needed.

Cloud computing platform is the internal support of cloud Computing center, at the core of cloud computing technology system. It is data-centric, with virtualization and scheduling technology as the means, by establishing a physical, scalable, configurable, bindable pool of computing resources, integrating server clusters and storage groups distributed across the network, and providing a secure and reliable variety of application data services, combined with the ability to dynamically allocate and smooth the expansion of resources.

2, Cloud Computing data Center implementation process

Cloud Computing Data Center implementation is not a simple software and hardware integration project, before implementation needs careful assessment and overall planning, fully consider the cloud Computing data center management model, and the future operation model into the overall planning, so that can fully play the role of cloud computing platform.

Combined with the survey of Cloud computing Center user needs and the implementation experience abroad, the current cloud computing data center infrastructure implementation is divided into the following 5 stages:

1 planning stage: to treat the cloud computing Center as a strategic issue, senior management should give great attention and support, and identify the goals to be achieved in each phase, from the Business Innovation and IT service transformation of the height of planning and deployment.

2 Preparation stage: According to the characteristics of the industry, fully understand the user using cloud Computing data Center to obtain the service and application requirements, and the cloud computing platform for adequate evaluation, select the appropriate technical framework. At the same time, fully consider the operability of system expansion and migration to ensure the continuity of infrastructure platform technology and core business continuity.

3 Implementation phase: Resource virtualization is the foundation of the Cloud Computing center, and it can meet the requirements of security, reliability, scalability, and flexibility by building a virtualized platform that supports heterogeneous platforms.

4 The deepening stage: on the basis of realizing platform architecture virtualization, we should also realize the automation of resource dispatching and distribution, and lay a good foundation for overall management and self-service service.

5 Application and Management phase: The basic feature of cloud computing is openness, and the cloud computing platform should be able to provide standard API implementations compatible with existing applications. All application porting is a step-by-step process, and the cloud infrastructure needs to be a good support for core applications, not just new requirements. At the same time, cloud computing platform construction is a closed-loop process that needs to be continuously improved.

3. Key technologies of cloud computing data center

The construction of the cloud Computing data Center incorporates a number of new technologies, mainly including the following.

1 Virtualization Technology

The application domain of virtualization technology involves many aspects, such as server, storage, network, application and desktop, and different types of virtualization technology solve different system performance problems from different angles.

Server virtualization is a rapid division and dynamic deployment of server resources, which reduces the complexity of the system, eliminates the disorderly spread of equipment, and achieves the goal of reducing operating costs and improving asset utilization.

Storage virtualization centralizes storage resources into a large pool of resources and is managed in a unified way, changing storage systems and data migrations without disruption and improving the dynamic adaptability of the entire system.

Network virtualization increases the number of connections and reduces network complexity by virtualizing a physical network node into multiple nodes and integrating multiple switches into a single virtual switch to optimize the capacity of the network.

Application virtualization helps improve service delivery by dynamically allocating resources to where they are most needed, and increases the availability and performance of applications.

Cloud computing Data centers implement system-wide virtualization across it architectures based on the virtualization technology described above, unified management, provisioning, and monitoring of all resources, simple and effective integration of a large number of decentralized, underutilized physical resources into a single large virtual resource without the need to extend important physical resources and enable it to run efficiently for a long time, thus maximizing energy efficiency and resource utilization.

2 flexible scaling and dynamic deployment

Elastic scaling can be considered in two aspects, vertical and horizontal. Vertical scalability means adding resources to the same logical unit to improve processing capabilities, such as increasing the CPU on an existing server or adding a hard disk to existing Raid/san storage; Horizontal scalability refers to adding more logical units of resources and consolidating them as a unit at work.

Dynamic provisioning is to allocate and manage the computing resources automatically according to the change of the requirement, and realize the scaling and optimization of the High "elasticity", while the users are not involved in the specific operation process.

3 Efficient and reliable data interchange and event handling

Data transmission Exchange and event processing system is the message and data interchange hub of cloud computing center, can not only use multicast protocol to pursue speed, and can not only use TCP to pursue reliability, but need to combine the advantages of a variety of protocols, effectively control the distribution of data flow between the various components of the network, to ensure the flow of data channels, The reliability and security of information exchange. At the same time, in order to meet the diversity of system application and the requirement of business real time, we should also consider many kinds of connection methods, such as point-to-point, point-to-point, multi point and multiple points.

4 storage, processing and access of massive data

Distributed mass data storage System includes two subsystems of distributed database and distributed file storage, which are used to deal with structured and unstructured data, as well as a series of adaptive tools compatible with traditional database and storage products to ensure the storage, access, synchronization and real-time migration, replication and Backup functions.

5 Intelligent management Monitoring and "plug-and-play" Deployment applications

Intelligent Management Monitoring System combines event-driven and cooperative cooperation mechanism to realize automatic intelligent management of large-scale computer cluster, it is not only responsible for automatically deploying, automatic upgrading, automatic configuration, visual management and real-time state monitoring for all the software services running on all servers. It will also be dynamically scheduled and automatically migrated according to the change of environment and requirement or the appearance of anomaly. At the same time at the system level for the entire distributed cluster of each component, regardless of hardware or software, real real-time, fully automated "Plug and Play" management.

6) Parallel Computing framework

The parallel computing framework is based on the integration of large-scale server clusters, combined with the design of a complete grid computing framework to ensure the collaboration between different nodes and individual nodes, the decentralized IT infrastructure is integrated together in a structured way to achieve high reliability, high performance and powerful data processing and computational analysis capabilities. The system automatically arranges and processes the complex work required to support distributed computing according to the task requirements and related data, and solves complex IT problems that require strong computing power such as business intelligence, business analysis, log analysis, and so on.

7 multiple leasing and on-demand billing

Multi-tenancy is the use of SLA setup means, according to the actual business characteristics and requirements, the performance and security of the whole system are optimized by the custom policy, so that the resources provided by the system can be processed on different granularity to form the characteristic service for different users, different use purposes and various performance forms.

On-demand billing is based on the unified allocation of resources, through monitoring and management mechanism to maintain the use of user resources tracking and recording and real-time feedback, in order to achieve user resources to use the measurement and service payment, save a lot of construction costs and operation and maintenance management costs.

4, Cloud Computing Data Center construction Supervision work key points

As the supervision of information engineering, it is important to consider cloud computing platform infrastructure system, product selection, technology realization and evaluation index system in the process of construction project of cloud computing data center. We should not only follow the implementation process, but also play the role of information consultant, and give the corresponding acceptance standards and norms to the key indexes such as resource running efficiency, monitoring management mechanism and service delivery process of cloud computing data center.

Supervision in the Cloud Computing data Center construction project should focus on the content:

1 Transformation of the infrastructure system

Cloud computing data centers differ significantly from the infrastructure architecture of traditional data centers, it may be a consolidation of multiple data centers, forming a unified cloud computing hub, and the internal data center is not just a simple integration of independent physical hardware and software systems, but a unified, tiered, virtualized resource pool of components. achieve higher utilization of equipment resources. At the same time, using service-oriented architecture, the application of different functional units (services) in a unified and common way to interact, so that the use and management of information systems more refined.

2 Product changes

Cloud Computing Data Center product composition is not only network equipment, servers, storage systems, operating systems, middleware and so on, but also need various types of platform virtualization software, automated migration/scheduling system, resource monitoring/distribution system, metering/billing system, as well as the service system and other related products, At the same time part of the function also needs to be customized development.

3 Integration of various technologies

Cloud computing Data center construction involves cloud network, virtualization, grid technology, parallel computing, storage technology, resource monitoring/metering, automated deployment/migration and self Service portal and other technical fields, according to the traditional integrated implementation of the project supervision work, it is difficult to grasp the overall completion of the project and application results, It is necessary to fully understand the technology used in the project in order to achieve the objective of substantive supervision.

To sum up, as long as we fully understand the cloud computing data center construction ideas, the overall framework, key technologies and core products, you can at all stages of project construction, control the main construction content and indicators, play the role of substantive supervision, supporting the construction party to achieve the project objectives.