Cloud applications have to be changed.

In our understanding, the cloud is a Web service that uses loosely coupled structures. As long as the host that provides the cloud serves on the internet, the user obtains the application and the service directly from the Internet server cluster. From this. As long as developers focus on the application of innovation and business logic, regardless of the operating system and hardware such as the underlying architecture, developers must also pay attention to the new challenges of cloud services.

By using APIs, developers can effectively write programs for cloud applications, and direct and efficient cloud providers require more resources. Also, developers need redundant design during the development process. If not, this can cause a server error.

Creating a cloud also requires designing a stateless cloud application. Stateless means that both the client and the server do not save each other's details, and the server handles only the current request without having to understand historical information.

The characteristic of cloud service is that if some applications fail, the application can be aborted and restarted completely. The stateless nature of the cloud reduces the amount of tasks that the server recovers from local errors. Cloud computing technology includes mechanisms to maintain application state changes, of course this requires stateless design. In the cloud, there is neither the concept of a local disk nor a real login, and all applications are stateless.

Simple cloud applications should be stateless. However, there are some interesting applications that require the retention of certain States, and therefore use a database to store some state information. Some specific projects, such as user information, can be stored, but applications with a certain size remain stateless in the cloud.

In the past, people generally developed the entire application on the local server, and now it's not necessary to do so. Another difference in cloud applications is that application components can run in different clouds. For example, an application can call Facebook's presentation layer and store data on Amazon's S3, where logic can be done elsewhere. This means that building cloud applications requires a new system architecture and a way of thinking, as well as the scale of cloud services.

The abstraction and statelessness of the cloud affect the database. For example, the Azure platform presents a model of a non-standard relational database, and the storage engine does not use a standard relational database. Therefore, a lot of work is meaningless if the application is developed according to the standard relational database. In addition, Azure's storage engine differs from Microsoft's cloud-based version of SQL Server, which developers must be aware of when designing.

At the moment, Microsoft is accelerating the development of Azure's ability to leverage relational databases, as well as providing data-table storage choices for relational data storage. The big Table database model used in APP engine is another example. "Big table is not a SQL database. Some of the features it supports are difficult to segment and cannot be combined with the idea of storing data across machines. "This means that developers upload code, and then Google manages and splits the database," said Pater Kummen, Google's App engine product manager. Such data processing has a specific pattern that is completely different from the traditional SQL model that developers must follow. ”

Cloud environments Discard relational databases. "In a high-traffic network environment, a relational database is difficult to manage and ultimately a huge burden for developers." "Sent Tes Kummen said.

"Edit Recommendation"

Technicians develop cloud computing from here. A new generation of open source cloud computing development platform to help learn cloud computing development of several open source platform "responsible editor: Mulberry Hill TEL: (010) 68476606"