A Brief Introduction to cloud computing

Original text from [than the Internet], reproduced Please retain the original link: http://storage.chinabyte.com/246/12169746.shtml

Cloud computing is becoming increasingly popular, and many users have begun to adopt this technology. I believe many readers will want to understand the principles of cloud computing. This article will introduce it in detail. Hope that the majority of readers like it.

Principle

Cloud computing is the development of distributed processing (Distributed Computing), parallel processing (Parallel Computing), and grid computing, or the commercial implementation of these computer science concepts.

The basic principle of cloud computing is that by distributing computing on a large number of Distributed Computers, rather than local computers or remote servers, enterprise data centers will run more like the Internet. This allows enterprises to switch resources to the desired applications and access computers and storage systems as needed.

This is a revolutionary action. For example, it is like switching from an old single generator model to a centralized power supply model for power plants. It means that the computing power can also be circulated as a commodity, just like gas, water, electricity, convenient access, and low cost. The biggest difference is that it is transmitted over the Internet.

The cloud computing blueprint is ready: in the future, we only need a notebook or a mobile phone to implement what we need through network services, and even tasks such as super computing. From this perspective, the end user is the real owner of cloud computing.

Cloud computing applications include the idea of combining power for each of the members.

Determine whether there are three criteria for cloud computing. What is cloud computing?

What is cloud computing? This is an issue that everyone is concerned about. Now we have found that there are many different ideas about what cloud is and what cloud is. Some people say that public clouds are clouds, while private clouds are not clouds. Others say that virtual clouds are supported, but virtual clouds are not supported, but Google does not, we all think that Google is a cloud. Some people say that 1000 servers are clouds, as if 999 servers are not clouds. Now there are some high-performance computing centers that haven't changed anything, that is, the names are changed to cloud computing centers. Why are there so many people playing the cloud computing flag? You can either make an eye or make money.

As the public, you need to identify which are real clouds and false clouds. To this end, Professor Liu Peng proposed three criteria for determining whether cloud computing is applicable for your reference:

The resources required by the user are not on the client and come from the network.

This is the fundamental concept of cloud computing, that is, providing the computing power, storage space, software functions and information services required by users through the network.

Service capability can be scaled in minutes or seconds

If the service capability of the resource node is insufficient but the network traffic comes up, it takes the platform to automatically and dynamically increase the number of service nodes within one minute, from 100 nodes to 150 nodes. Cloud computing requires enough resources to cope with the peak traffic of the network, even if it suddenly becomes a hot spot like "Jia Junpeng. After a while, the traffic goes down, and the number of service nodes decreases as the traffic decreases. Some traditional IDCs claim to be able to provide scaling capabilities, but it takes more than an hour to provide them to users. The problem is that network traffic is unpredictable and cannot be waited for so long.

Compared with the traditional model, it has a performance-price advantage of more than 5 times.

After reading the above article, some people are thinking, it doesn't matter. If you have more machines, the traffic will be heavy enough. But this is not the concept of cloud computing. We also have a performance-price index. Cloud computing is an epoch-making technology because it puts a large number of cheap computers into the resource pool and uses software fault tolerance to reduce hardware costs, by deploying cloud computing facilities in cold and abundant electric resources, you can save power costs and improve resource utilization through large-scale sharing.

Representative cloud computing platform providers in foreign countries have achieved an astonishing 10-40 times the performance-price increase. Due to technical, scale, and unified electricity price problems in China, it is difficult to achieve the same performance-price ratio for the moment. We have set this indicator to five times. The cloud computing platform of China Mobile Research Institute, with 256 nodes, has increased the performance-price ratio by 5 to 7 times. The performance-price ratio also has room for improvement as the scale and utilization rate increase.

Example

These three standards are equivalent to three filters, which can filter out things that are not in the cloud computing category. Let's try to see if these are cloud computing:

PC system: the first step is not to pass, because all the resources required by the user are on the client, so they are not in the cloud computing category;

IPhone application software: If the app can run independently after it is downloaded to the iPhone and only communicates with the outside world, it cannot go through the first level. If the app runs on a network platform, computing and data resources come from the network. If the iPhone is only an access interface, it may enter the cloud computing field;

Web site: has passed the first level. However, if the scalability level is reached, it will be hard to understand the performance-price ratio;

Netbook: it can pass the first level, but it is only a terminal. If it is only used for Internet access, it can only be included in the cloud computing category if it is used for interaction with cloud computing facilities.

AD Alliance: Second level. By combining the advertisements of different websites, an advertising alliance is formed. An advertisement can be published on many websites. However, because the public does not have a high click rate on ads, AD alliances do not need to have a large amount of scalability, and it is not easy to do it-because the participants are a large number of small websites;

Super Computing Center: if the scale is large enough, the second level is barely passed, but the cost effectiveness is not good. Super Computers are expensive and have no advantage in cost-effectiveness. We do not think it is cloud;

Online office: it is easy to meet cloud computing requirements after large-scale operation. For example, we can see Baihui office and Google Docs today.

Online CRM: it is easy to meet cloud computing requirements after large-scale operation. The cost of providing users with a month is tens of dollars, which is far better than the traditional mode, which is cloud;

IAAs: like Amazon's service for renting machines, this is also very cost-effective. It only costs $0.1 to rent a virtual machine for an hour. It is also cloud.

PAAs: for Microsoft's azure, 25 GB of storage space is provided free of charge, which must be very cost-effective and cloud-based.

Cloud security: after large-scale operation, it is easy to meet cloud computing requirements. For example, 360 security guard provides users with completely free services. This is because of high cost effectiveness.

Original text from [than the Internet], reproduced Please retain the original link: http://storage.chinabyte.com/246/12169746.shtml