-
Book Overview & Buying
-
Table Of Contents
Integrating Hadoop
By :
ETL (extract, transform, load) and ELT (extract, load, transform) are both acronyms for three-step processes that move data from one place (and purpose) to another. Generally, data from multiple source systems is being moved to, and consolidated in, an enterprise data warehouse (EDW) or other target database(s), where it becomes available for further use.
The important difference between ETL and ELT is in the transformation step. In this step, data is cleansed, put into formats/structures required by the EDW or downstream datamarts and apps, and normalized and integrated so that it can be compared, merged, and analyzed with data from other sources. When using ETL, these tasks are performed through automated tools or hand-coded scripts prior to loading. With ELT, the bulk of transformation is completed after loading, inside the HDFS, data warehouse, or other target database. Pre-transformations can take place in the source database as well.
Change the font size
Change margin width
Change background colour