How do I map an application to a mixed cloud?

There is little doubt that mapping the right application to a mixed cloud of the enterprise can determine how much value it can recover from the public cloud. Most enterprise IT applications can be grouped into the following categories, each of which has its own problems:

• Trading Business Network applications: for users connected to the network, provide login services, customer support, or other business-oriented activities.

• Private Business applications: Provides access to corporate applications, but does not provide general access to the public or to customers.

SOA Distributed Application: Based on the service-oriented Architecture component model, you can run one part of the application in a distributed manner.

• Data-mining applications: accessing historical data to enhance business intelligence (BI).

Trading Business Network applications not only support customers or vendors, but also collaborate with network front-end applications in data centers and enable remote access. Transaction Business Network Cloud computing model is the most common application model deployed in the data center.

It is now the practice of the industry to host these applications using a public cloud computing service. Then use a VPN link to transfer the business from the network server to an application server located in the corporate data center.

Enterprise data is still in the data center, so information security and data storage costs are low. The only significant security and consistency problem you may encounter is ensuring the security of network hosted application server connections, and uninstalling network hosting can improve the use and cost of data center resources. If you're looking for a hybrid cloud, that's a good start.

A private business application is a problematic application model because it increases the location of data storage. The cost of public cloud computing data storage varies enormously; some can be very high, and a single month's cost equals the cost of a disk drive with the same capacity. For an enterprise where storage costs are an important consideration, it is best to store the data in a data center and then use a remote SQL Server query to access it.

The private business model has its use value in the scene where the trading business needs the user to access a large amount of data. For security reasons, most companies that encounter similar problems look for cloud providers that integrate configured VPN services with cloud computing products. A configured VPN allows its enterprise to obtain a meaningful application service level agreement (SLA) that is not available to the Internet.

If applications hosted in the cloud need access to the enterprise's Data Warehouse, SOA distributed applications increase the cost of data storage; Remote SQL Server queries can compensate for this. But application components distributed through SOA can create additional security and performance problems, especially if you use the Internet to connect enterprise users, Enterprise managed SOA components, and cloud hosting components.

In this case, the key is to have a private VPN connection through a network vendor with a strict SLA. Even when accessing the cloud through a VPN, you must check how the application component is managed to ensure that the work is not replicated between the data center and the cloud. Because duplication of data can lead to delays and even negative effects on normal production.

Of the four types of enterprise IT applications we discuss, data mining applications will produce the most expensive data storage costs. Still, the unpredictability of BI and data mining applications makes them the most interesting candidate for two cloud computing resources. It is also possible to allow data mining based on cloud computing or bi-type applications to access data centers through remote SQL queries, but the amount of information needed to transmit data over a network can cause performance problems.

The best way to migrate data mining-type originality to a mixed cloud is to create an abstract database that summarizes a large number of enterprise collection history data. You can then store the aggregated database in the cloud and access it through the public cloud BI tool without paying expensive fees. In addition, maintaining detailed core data in the cloud is risky.

Original address: http://www.searchcloudcomputing.com.cn/showcontent_52656.htm