Hybrid data Warehouse and the new structure of the BIG Data Warehouse

(Reading notes)
Many companies, even if they want to bring Big Data, still need to continue to use Data Warehouse to manage the structured statistics and system records. and Big data, for data Warehouse, provides a complementary opportunity, not a substitute for the latter.

Highly structured information (data) can still be retained in data Warehouse, while decentralized (distributed) information, as well as changes in time, can be controlled by the Hadoop-based structure of the material.

Figure 1 Traditional data Warehouse and data Mart structures

Figure 2 Hybrid Data Warehouse and the new structure of the BIG Data Warehouse

A company's customers, their potential users, their interactions on the web and on the physical, these massive Data Warehouse can only track transactions and traffic, but they can't keep track of what's going on and what's on the Internet. As much as it is possible to keep these Data, and to set up a digital Warehouse to store all of them, it is better to use the Hadoop decentralized approach to storing the numbers on the company's servers. In this way, the company will be able to save all of the "web interaction" of the digital. These are stored in the jungle set of servers for Hadoop and MapReduce (cluster), and with tools like Flume and Sqoop, the company's information team can move the numbers out of the Hadoop climate into the relational model and repositories, so that you get a familiar The traditional SQL tools to check.

This way, companies can quickly transform services and products as they discover certain customer groups in the hope of gaining some new forms of service. The company (online travel website) can also predict some trends, such as how to adjust the price of the ticket. Some of these data remain in the HADOOP environment and can maintain almost "instant" updates, while others have been transferred to Data Warehouse, so they can be used in comparison with historical figures in the past. "------yes. The existing Data Warehouse continues to provide the content required by the company's business, and the Hadoop environment can keep track of what happens every minute. This dynamic large Data system, which integrates the system record with the Warehouse, can provide a huge business opportunity for companies to use the company's business in the web world to produce a large amount of time-generated data and analysis results.

----------------------------------------------
The above is recorded from Big Data for Dummies simplified
Ch11, equipment and large repositories

The book 260 pages, more graphics, the content of the principle, technology, and enterprise integration of existing applications, almost no code (suitable for the boss and supervisor to see)

Big data for everyone (Simplified translator):
http://www.m.sanmin.com.tw/Product/Index/004706578
isbn13:9787115356130
isbn:9781118504222

Hybrid data Warehouse and the new structure of the Big Data Warehouse