The business model of cloud computing must find a new "landing point"

Through analysis and observation, we found that the current cloud computing is mostly personal computing business model, for enterprise computing we can see is the data center (ie IAAS/IDC), which is based on IT resources a business model. But in essence, enterprise informatization needs to solve the problem of application complexity, a cloud computing model based on application-driven. Before we show that cloud computing is not enough to meet the needs of enterprise Informatization, let's take a look at how the Saas/paas/iaas we preach has eclipsed it.

First, analyze the concept of SaaS, which is more familiar than cloud computing. At present, many people will be the reason for the failure of SaaS promotion, blamed on companies do not trust the data in the cloud provider. But just as people do not trust the money in the bank rather than under the pillow, and then slowly accepted the bank's preservation and treatment. Worrying about data security is a superficial phenomenon, it's just a feeling of insecurity that sometimes doesn't necessarily depend on it. Imagine that if SaaS solves such a security concern, there are several deeper issues to be addressed in order for SaaS to be truly widely used in the enterprise.

The first thing to mention is personalization. The need for personalized enterprise applications will only become more and more strong, but without scalability and personalized customization features of the SaaS service, for a small size of the enterprise applicability is a very big problem.

Second, the service migration problem of cloud computing. The nature of cloud computing is Service computing, service computing and the sale of software as an asset to the enterprise is very different, that is, service continuity and service migration issues. How does the service migrate? What about the enterprise's data if the enterprise doesn't need this SaaS service today? The data is an enterprise, legally speaking, the enterprise does not have the assets of SaaS Software, but the data generated by SaaS is the enterprise's assets, after the Stop SaaS service, how to retrieve the data assets belonging to the enterprise? There is no SaaS application, this data is not useful to you, how do you query, how to deal with? service continuity and service migration allows cloud service providers to increase customer stickiness and achieve user lock-in, which is advantageous to cloud service providers, but limits the user's choice of permissions.

Third, interoperability issues with cloud services. None of the SaaS vendors have been able to provide all the business services to the enterprise so far, and not in the future. This presents a problem where an enterprise may require application services from multiple SaaS providers, so how does the interoperability between cloud services work, and how are the integrations between these cloud services resolved? If we assume that the SaaS model is a future trend, there must be a "override" in addition to SaaS An operating service provider above the SaaS provider that provides services or platforms for the interconnection of clouds and clouds. As the network connected to the network is called the Internet, there is an interconnected cloud (intercloud) between the cloud and the cloud. So far, there are no interconnected cloud technologies, concepts, and solutions.

Therefore, since SaaS does not yet have the means to provide the services required by large enterprises, the conditions for large enterprises to use SaaS are premature. SMEs do not use the reason is because Chinese SMEs are still in an immature market mechanism conditions, their survival conditions and market environment is very bad, still need time.

Second, let's look at the PAAs problem. Now all internet companies are turning to what is called an open platform, careful analysis, you can see by Google, Tencent and other proposed open cloud computing development platform, is actually based on private architecture and personal application-oriented two-time development interface, the development of language is dedicated, running platform does not support the mainstream. NET or Java EE architecture, can not be called a real PAAs, support enterprise ERP run development is difficult to achieve. Essentially, their so-called open platform is not an "open" platform, at best, just a "let go" platform, open to some partners can be based on their infrastructure to expand a part of the new application, and the development of a new business process-driven enterprise application system is basically not feasible. As a result, there is a long way to go before all the declared PAAs platforms support enterprise applications, and the market needs a truly open PAAs platform.

Finally, look at IaaS. IaaS is one of the lowest levels of our three tiered models of cloud computing. IaaS claims to cost savings, including hardware and related investment costs, saving electricity, renting hardware servers (cheaper than buying), and not building a computer room. But through analysis we can get, enterprise informatization cost inside hardware purchase cost and electric power consumption in the whole informationization investment, only accounted for a very small proportion (as shown below). The biggest investment in enterprise Informatization is in the aspects of operation, training, management, service and so on, but there is no solution to the problem.

In the enterprise's cost composition, including the explicit cost and the recessive cost two aspects. We should not look only at the visible cost of those inherent hardware inputs, but also consider the hidden cost of the business in the process of operation. Gartner's research also indicates that CIOs use cloud computing to reduce costs, and must focus more on data, applications, and consolidation than on obvious.

To sum up the analysis, at least for now, the cloud of the three models Saas/paas/iaas in the field of enterprise information is not workable, rather than the outside show a thriving. Cloud computing business model to meet the needs of enterprise informatization, must find a new landing point.

(Responsible editor: Lu Guang)