How to build a killer application with large data

We have not given enough attention to the growing boom in big data. While large data can provide stunning business insights, it is worthless unless these business insights are presented in an application that inspires new business practices.

Virtual machines have been a http://www.aliyun.com/zixun/aggregation/14385.html in the field of "> Cloud Applications" for a while, and now my mom should turn her attention to the intersection of big data and cloud computing.

How do you develop an application that can easily move through private and public clouds and receive data on both sides of the firewall? So how do you build big-data killer cloud applications? Here are some of our initial ideas:

1. Lightweight services: Rest is the new soa--lightweight service that forms the basis for supporting Web front-end, and publish/subscribe messages like RABBITMQ form the basis for back-end workflows and processes.

2. Mobile-Priority user interface: The Twitter boot library allows developers to develop HTML5 applications for mobile devices, thereby landing on flat-panel and browser-based computers.

3. Fast data: Extending the front-end of an application often requires data management in memory. The easiest way to interact with core data is through a SQL interface, such as Sqlfire.

4. Big data: Knowing what's going on now requires fast data and a lot of historical data to know what to do. The key here is to provide consolidation between two data sources to keep the data warehouse and memory database data as fast as possible.

5. Application-tier management: the performance of the management application as a series of logical tiers, not physical instances, greatly reduces the administrative complexity of the system administrator.

6. Cloud Deployment: Automated, development/operation solutions like creator Director, removing black magic from large-scale system deployments, reducing the number of days of deployment to a few minutes of script.

7. Flex expansion: The core value of a cloud computing environment like cloud Foundry is to keep the size of the task computing resources in hand--when demand is high, the size of the resource expands, and vice versa.

Self-healing: The cloud means you never have to say you're sorry when your site crashes because of a component and it can't be restarted-again, Cloud Foundry comes in handy. (Translator/Lin Xu, words)

(Responsible editor: The good of the Legacy)