Why Data Warehouse
Why Data Warehouse
Why Data Warehouse
why we go for data warehouse?
Why a Data Warehouse?
It allows for the following benefits to the business and Business Intelligence layer:
1.Speed: The data structure in a Data Warehouse (Star Schema) is the optimal structure for delivering a result set to the Business Intelligence layer; and hence is many times faster then going against Database directly, which has a structure optimized for storing data, not delivering it
2.Pre-Gathered Values: One of the primary challenges in producing reports against Database directly is developing the logical joins and relationships in the data to appropriately deliver reports. In a Data Warehouse these values are all prepared up front. This leads to benefit #3
3.Development Speed: If a new report needs to be created against Database today, it will require piecing together the different tables in the database. Having this work done up front allows reports and universe elements to be created quickly
4.Offloaded Processing: When the Business Intelligence layer requests data from the database, the database will require processing power to produce the result set. If Database is being hit directly with this request it will take processing power away from other resources that need the Database database. Namely the applications themselves
5.Independence: If the organization ever decides to move away from their application (say a CRM) to another application (another CRM) the Business Intelligence layer can remain in tact as the majority of changes would occur in the Data Warehouse. However, if Database is being accessed directly by the reports then the organization would be required to rewrite the logic for each and every report individually
6.Singularity: The Data Warehouse delivers a single location from which the data is extracted giving confidence to the business that there is a single version of the truth. Additionally, IT benefits from a higher ROI per hour spent producing data relationship logic (in the Data Warehouse), as the logic can be leveraged over and over by the Business Intelligence layer
Important things to be aware of in a Data Warehouse
The benefits of a Data Warehouse are great but they require and awareness of the following items:
1.Scheduled Data: The data in a Data Warehouse is extracted (from Database), transformed, and then loaded into the Data Warehouse (another database). This is a scheduled process and thus the data will not be “live” against the database. This doesn’t mean that values cannot be pulled live for certain report values outside the Data Warehouse; but in general the Business Intelligence implementation will be pulling from data that is a day or half-day old. This can also become a benefit, as the schedules can be kicked off when other processes finish, thus avoiding the danger of delivering incomplete numbers to the business. (ie. receiving a report that ran before the accounting books closed in the system, producing incompatible results, unbeknownst to the business)
2.Investment: The implementation of the Data Warehouse needs to be seen as a long term investment. Inevitably, down the road there will be business processes that change, or business that are added, requiring these results to be reflected in the Data Warehouse. Hence, best practices, and a strong starting point are critical.
Why Integrate Hadoop into a Data Warehouse? (AD)
