Analysis: Six key technologies in the process of cloud computing development

Cloud computing is data-centric, data-intensive supercomputing. It has its own unique technology in data storage, data management, programming mode, concurrency control and system management.

Six key technologies of cloud computing

Mass Distributed storage technology

To ensure high availability, high reliability and economy, cloud computing uses distributed storage to store data and redundant storage to ensure the reliability of stored data, and a highly reliable software to Ibuyingjian, and thus provide low-cost, reliable systems. In order to meet the needs of a large number of users, data storage technology must be characterized by high throughput and high transmission rate.

The cloud's data storage system is mainly the Open-source system HDFs (HADOP Distributed File system) of Google GFS (Google File system) and the Hadoop development team. Most it vendors, including Yahoo and Intel's "cloud" program, adopt HDFS data storage technologies.

Parallel Programming Mode

In order to efficiently utilize cloud computing resources to make it easier for users to enjoy the services that cloud computing brings, cloud computing must ensure that complex parallel execution and task scheduling in the background are transparent to users and programmers. Cloud computing uses MapReduce programming mode, which automatically divides tasks into subtasks and dispatches and assigns tasks to large scale computing nodes through map and reduce two parts.

Data management Technology

Cloud computing system for large data sets processing, analysis, to provide users with efficient services. Therefore, data management technology must be able to efficiently manage large datasets. Secondly, how to find specific data in large scale data is also the problem that cloud computing data management technology must solve. The data management of the cloud system often adopts the data managing mode of the column storage, which guarantees the performance of massive data storage and analysis. The most famous data management technology of cloud computing is Googe's bigtable data management technology, and colleague Hadop development team developed the BigTable Open source data management module HBASE.

Distributed Resource management technology

Distributed resource management system is the key technology to ensure the correctness of system State in the Multi node concurrent execution environment. The system state needs to be synchronized between multiple nodes, the key nodes need to migrate service, and the distributed resource management technology can coordinate the use of resources by locking mechanism to ensure the consistency of data operation. Google's ghubby is the most cone of your distributed resource management system.

Cloud computing Platform Management technology

Cloud computing resources are huge, and a system can have as many as 100,000 servers, spanning several data centers located in different physical locations, and running hundreds of applications. How to manage these servers effectively ensures that the systems that make up these servers provide 7 * 24-hour uninterrupted service is a huge challenge. Cloud computing System Management technology is the "neural network" of cloud computing, through these technologies can be a large number of servers working together to facilitate the deployment and opening of the business, rapid detection and recovery of system failures, through automated, intelligent means to achieve the assembly mill system can operate, manageable. Google, through its superior cloud computing management system, maintains the world's millions of PC servers in a coordinated, efficient operation, and cloud computing Systems management technology is also the core of corporate confidentiality, so far no release of any technical information.

Green Energy-saving technology

Cloud computing technology reduces the cost of purchasing the server, which makes the operating cost of power consumption become one of the main expenses of computing center, in order to further reduce costs, cloud computing first to the green energy-saving technology in a lot of exploration.

The cost of air conditioning cooling in traditional data centers is about the 30%-70%,google of the overall electricity bill in order to save energy, the cooling system in the data center in Belgium does not need coolant, but takes the outdoor air for natural cooling. According to the Belgian climatic conditions, the free cooling mode is not available for only 7 days a year. In the hot summer, Google can transfer the computing tasks of the data center to other data centers through the cloud management system.

(Responsible editor: The good of the Legacy)