The widespread realization of cloud computing is the IT world of Communism, "dens, on demand". To 2011, cloud computing in the domestic development of the key words from the "concept hype", "technical discussion" into "Application practice." The major IT vendors are actively Shinma, providing enterprise users with a variety of product solution options. But what does cloud computing mean for a specific domestic enterprise CIO? What stage of cloud computing is he in now? How to evolve to cloud computing? These issues are as relevant as the theory of social primary stage.
Enterprise's It construction process, measured by the current benchmark, to cloud computing evolution, basically can be divided into three stages, as shown in the figure.
First stage: Large concentration process
The process of the enterprise decentralized data resources, it resources of the physical concentration, forming a large-scale data center infrastructure. During a dataset, continuous implementation of data and business integration, most of the data center of the enterprise has basically completed its own standardization, so that both the expansion of the business and new business deployment can be planned, controllable, and enterprise standards for IT business implementation, to solve the data business dispersed period of chaos and disorder problem.
In this stage, many enterprises in the late data concentration began to disaster tolerance construction, especially in the snowstorm, after the earthquake, the enterprise disaster-tolerant center construction generally received attention to finance as the hot industry almost carried out the industry-wide disaster tolerance construction boom, and the financial industry most disaster-tolerant construction of the level are very high, Target application-level disaster (data 0 loss).
In general, the first phase of the process to solve the enterprise IT decentralized management and disaster recovery problems.
The second phase: the process of implementing virtualization
After data concentration and disaster tolerance are realized, with the rapid development of the enterprise, the data center IT infrastructure expands quickly, but the system construction cost is high, the cycle is long, even if is the standardized business module construction (even if is the system duplication construction), the hardware and software purchase cost, the debugging running cost and the business realization cycle did not drop significantly. Standardization does not bring flexibility to the system, centralized large-scale IT infrastructure, there is a large number of system utilization problems, different systems running in exclusive hardware resources, inefficient and data center energy consumption, space problems gradually highlighted. As a result, virtualization, which aims to reduce costs, increase the flexibility of IT operations, and increase resource utilization, is beginning to be deployed in the data center.
Virtualization masks the heterogeneity of different physical devices, virtualization of physical resources based on standardized interfaces into logical computing resources (virtual machines) and logical storage space that are logically fully standardized and consistent. Virtualization can consolidate multiple physical servers into a single platform, each server runs a variety of applications of virtual machines, to achieve the increase in the utilization of physical server resources, because the virtualized environment can achieve the calculation and storage resources of the logical changes, especially the virtual machine cloning, so that the data center it implementation flexibility greatly increased, The business deployment cycle can be narrowed down to one day within a few months. After virtualization, applications run on a VM as a unit, and the number of data center servers can be greatly reduced and computing energy efficiency improved, resulting in the control of energy and space problems in the data center.
In general, the second phase process enhances the flexibility of the enterprise IT architecture, the data Center resource utilization efficiency increases, the running cost is reduced.
Phase three: Cloud computing phase
For businesses, data center systems (including hardware and software and infrastructure) are a large resource input. The new system (especially the hardware) after the completion of the general experience of 3-5 years to face the gradual aging and replacement, and software technology is constantly facing the pressure of escalation. On the other hand, it's input is difficult to match the needs of the business, even if the virtualization, it is difficult to solve the increasing demand for resources, in a certain period of scalability is always limited.
So enterprise IT produces a new blueprint of expectations: IT resources can be flexible expansion, on-demand services, the service as the core of it, improve business agility, and further reduce costs. As a result, service-oriented It requirements began to evolve into cloud computing architectures. Cloud computing architecture can be built by the enterprise itself, third-party cloud facilities can also be used, but the basic trend is that enterprises will gradually take the way to hire IT resources to achieve business needs, like hydraulic, power resources, computing, storage, network will become the enterprise IT operation of a use of resources, without their own construction, can be obtained on demand. From a corporate perspective, cloud computing addresses the dynamic and final cost of IT resources, enabling IT departments to focus on service delivery and business operations.
In the above three stages, the large concentration and disaster tolerance are the data center physical component and the business module, virtualization is the data center-oriented computing and storage resources, cloud computing is ultimately oriented to IT services. Such an evolutionary process shows a gradual change in the IT operations model, cloud computing, in turn, fundamentally changes the traditional IT service structure, stripping away factors unrelated to the core business of the IT system (such as IT infrastructure), integrating it with the core business, and adapting the enterprise's IT service capabilities to the changes in its business.
During the three phases of it practical construction, network is the bottom of the pipeline, so usually the "network first" idea of construction, in the network architecture design to fully consider the evolution process will face challenges and solutions, to achieve the network layer of cloud ready to host the future of cloud computing it platform.
For the real evolution of the Enterprise cloud computing model, on the one hand need the enterprise CIO to understand the latest technology trends; On the other hand, and more importantly, the CIO needs to develop an evolutionary plan for enterprise cloud computing based on business characteristics, culture and management patterns, current IT resources and capabilities, Clear the focus of the current phase of construction. "The cloud is not a goal, it's just a means to perform calculations." No enterprise should abandon its accumulated it investments and start over again, and only follow a gradual, non-destructive evolution to make the cloud truly deployed.