A very renowned technical expert once said this: “For cloud computing, everyone has their own definition.” There is a consensus among people regarding private clouds. From the architectural perspective, cloud computing is divided into three layers: SAS, PASS, and LASS. There are also common classifications such as public clouds, private clouds, and hybrid clouds.
Cloud computing is the development of parallel computing, distributed computing and grid computing, or the commercial realization of these computer science concepts. Cloud computing is the result of the mixed evolution and leap of concepts such as virtualization, utility computing, laaS (Infrastructure as a Service), Paas (Platform as a Service), and SasS (Software as a Service).
1.Four significant features of cloud computing
1) Cloud computing offers the most reliable and secure data storage center, eliminating the worries of data loss and virus invasion for users.
2) Cloud computing has the lowest requirements for user-end devices and is the most convenient to use.
3) Cloud computing can easily achieve data and application sharing among different devices.
4) Cloud computing provides us with almost unlimited possibilities for using the network.
2.Intra-enterprise private cloud
2.1 Overview of Intra-Enterprise Private Cloud. A private cloud is built specifically for a single customer, thereby providing the most effective control over data, security, and service quality. The company owns the infrastructure and can control the way applications are deployed on this infrastructure. Private clouds can be deployed within the firewall of the enterprise data center, or they can be hosted in a secure hosting facility.
Private clouds can be built either by the company’s own IT department or by cloud providers. In this “managed dedicated” model, cloud computing providers such as Sun and IBM can install, configure and operate the infrastructure to support a dedicated cloud within a company’s enterprise data center. This model gives the company extremely high levels of control over the usage of cloud resources, while also providing the specialized knowledge required to establish and operate the environment.
2.2 Internal Private Cloud Model of the Enterprise. Figure 1 depicts an infrastructure-as-a-service cloud computing model. It shows three types of clouds. Enterprise internal private cloud A and Enterprise internal private cloud B, and the external cloud is the role played by the service provider.
It is not the case that every workstation is a cloud. On the contrary, each site offers an IT-as-a-service delivery model. The system is supported internally. Some small businesses might consider integrating the virtualization architecture management layer of the external cloud with the internal cloud, thereby deploying internal virtual machine image files and running them on the architecture of the external provider. This can provide support for business areas such as training and research and development.
In the server of the “cloud management” scheme shown in Figure 2, the task division is as follows:
The function of the external cloud is to manage the servers. There are two servers in total, one is the “primary” server and the other is the “backup” server. On these servers, the “heart” of the software runs, which is used for the predictive power engine. It is responsible for the prediction, analysis, scheduling and management of the entire cloud computing tasks.
In the left part of the above picture, there are 4 internal private cloud monitoring servers for the enterprise. They are responsible for collecting data from all monitoring items in the enterprise’s data center and network. After being “virtualized”, the computing capacity of these four servers can reach a maximum monitoring scale of 60,000 points (with a polling time of 2 minutes, meaning that all 60,000 points need to be polled once every 2 minutes). The average load on each monitoring server for these 60,000 points is 15,000 points. Both the monitoring capacity of a single server and the overall management capability reach world-leading levels.
On the right side of the above picture is the enterprise’s internal private cloud independent database, consisting of 2 units. All the monitoring, alerts, reports, management, views, and topological graphics and data are stored in this database. The greatest advantage of the independent database is the storage and management of massive data, based on the “Memory Cache” high-speed caching processing technology. Under the cloud computing solution with 4 monitoring servers, 60,000 monitoring points, and a long polling interval time, the independent database needs to process up to 5 gigabytes of data every day. The processing efficiency and speed of such massive data is unimaginable for traditional databases.
Enterprise-level private cloud must be autonomous, agile, and possess the flexible, efficient and secure characteristics of the cloud. In addition, enterprise-level private cloud also needs to have the following features:
The flexibility of private clouds enables them to adapt to unexpected situations, meet unplanned demands, and accelerate the configuration of new or constantly changing workloads, thereby enhancing the return on investment for projects.
Secondly, the efficiency of private cloud lies in its ability to reduce costs and align costs with IT/ business goals. Due to the low initial investment, scalable expansion, and advantages such as streamlined configuration, streamlined replication, and streamlined conversion of private cloud, users can purchase storage capacity as needed for new application data, thereby reducing capacity and related costs (SAN, energy, space, etc.) by up to 75%. Moreover, because private cloud can complete storage configuration in a few seconds and does not require pre-purchasing storage capacity, Shenduo can shorten the management time by 90%.
Finally, private clouds offer excellent security and eliminate the loss of shared assets. Storage technologies similar to management programs can ensure the security of dedicated assets of shared cost organizations and maintain the security of 100 virtual arrays in a single physical environment. The fully automated load balancing, as well as support for mixed loads and carrier availability, can ensure service levels. Therefore, in the event of a failure, the performance remains strong and stable.
3 Summary
The advantages of cloud computing have helped enterprises control cost management. Private cloud represents the convergence of enterprise computing technologies and is more powerful than modular general-purpose integrated hardware. It can be divided into several very small parts, and the network and storage can be dynamically allocated through the best strategies. However, as an emerging technology, it also has many shortcomings. For example, due to excessive traffic, the website of online retail giant Amazon also experienced abnormal access on “Black Friday”. And the EC2 “cloud computing” service promoted by Amazon was also launched to solve similar problems. Additionally, the “Software as a Service (SaaS)” provider Salesforce.com experienced a downtime of about 40 minutes a few days ago.
Some supporters of “cloud computing” technology claim that through this technology, enterprises’ storage space, bandwidth and computing power will be “permanently expanded”. However, in actual business operations, it is not so easy to achieve this. With the improvement and establishment of the cloud computing system structure, its technology will undoubtedly have a huge development space and broad application prospects in China.