There are too many articles on cloud computing online, this article mainly in the perspective of enterprise IT deployment of cloud computing to analyze the problem, let us first take a look at those feasible models. The application of cloud computing realizes the communism in the IT community, "doing everything we can and allocating as needed." This everyone got unanimous approval. Of course, for a specific domestic business CIO, cloud computing for their business in the end what it means. Cloud computing is currently in the cloud computing what stage it. What patterns of evolution do we see? These issues are more explored than those in life.
Before discussing the enterprise's cloud computing path in detail, it is necessary to conduct a physical examination of the state of the domestic cloud computing first. In accordance with the current common sense model, cloud computing stack has three levels, SaaS, PaaS, and IaaS.
SaaS model can be understood as the virtualization of service-based business functions, such as providing CRM on Demand services to multiple enterprises. Recently, the hot topic has been to include Web 2.0, such as social networks, Twitter, as a channel, into a marketing-oriented service Inside the cloud In China, SaaS has also made some progress in the past few years, but the type of business is very different from that in the United States. At present, 90% of domestic SaaS businesses are tool-based, such as email, antivirus, OA, financial software for small and medium-sized enterprises, and have not touched the core business of large-scale enterprise IT. In other words, the domestic SaaS business is cost-oriented in the first place. Some SMEs are reluctant to set up their own financial applications, mail systems and anti-virus systems. They reduce the cost by using the on-demand mode provided by the application vendors. On the application provider side , SaaS model provides a mode to the low-end market, both to avoid piracy, but also access to customers. Another benefit of the SaaS model is that it is theoretically possible to provide a development and usage platform based on the SaaS model to development providers such as small and medium-sized customers and software parks. Software sales, both in licensing and pay-as-you-go models, are essentially paid by some for code-based mental work, simply because they are intangibles and are different through technology channels.
SaaS model in the country have encountered two major challenges, the first is the security aspect, the second is manageable. Of course, the same applies to the cloud computing model. Putting production and trading data on someone else's machine and even delegating it to someone else requires a great deal of trust and assurance. In the short term, this is hard for most businesses to try, unless it is a tool-based application. SaaS model provides the software development of the market model, but also bring about multi-tenancy management, functional granularity, software reuse, etc. Think about it a headache topic.
PaaS model provides a model for the second problem, Amazon's practice is to provide a platform directly as a provider, the entire platform used to do their own e-commerce outsourced, it seems SaaS ideal many-to-many model too Multivariate, a platform leveraging multiple vendors by a robust technology platform vendor is more viable. This platform model provides two references for domestic IT. One is to provide a sector-specific platform, a similar regional medical service center and a regional education platform by the government departments to provide specific public services for specific small and medium-sized objects (hospitals and elementary schools) in the district such as patient health records, Student management services, etc., this way to solve the imbalance of IT resources. Another benefit of regional platformization is the combination of service and management.
PaaS model provides a choice of private cloud for large enterprises, especially for group enterprises. As shown in Figure 2, collective enterprises can build an internal private cloud platform through the stages of IT centralization, gridding and virtualization. Subsidiaries of various IT services, internal use SLA model. This model because of the occurrence within the enterprise, to avoid the security concerns, the use of a platform model, but also to avoid the challenges of service management.
IaaS model has a realistic simple example, a customer will move his server directly to the operator room, through the remote camera, the customer's IT maintenance staff can monitor their own server operation. IaaS model emphasizes the effective use of IT resources, in this respect, IaaS and IDC centers around the year 2000 there are similarities, the difference is that the current virtualization technology has been developed to a new level. In the application server, database server, storage virtualization, WAN data replication level, the technical level can reach a higher level. For example, an IT company composed of 2,600 units for software development to test the whole process of a private cloud for more than 20,000 engineers. Customers can now back up their data to a storage cloud and change the hardware partition of the database server at any time to support application clusters on thousands of PC servers. Multiple levels of virtualization are the foundation behind which the IaaS model is based. This is both a key requirement for Enterprise IT to build a private cloud and an operator base for IaaS for enterprise customers. Of course, the latter needs to address security concerns.
For the real evolution of the cloud computing model, on the one hand, enterprise CIOs need to know the latest technology trends. On the other hand, perhaps more importantly, CIOs need to make decisions based on business characteristics, culture and governance models, current IT resources and capabilities The convergence of elements to find a particular stage of the evolution path, the model of the private cloud for CIOs in large enterprises provides a possibility.