What is better, ETL or ELT?

ETL is the most common technique used when transferring data from a source system to a data warehouse. But there are cases where you might want to use ELT. What is the difference between the two? Read on:
Extract, Transform, and Load (ETL) is a process that involves extracting data from outside sources, transforming it to fit operational needs (sometimes using staging tables), then loading it into the end target database or data warehouse. This approach is reasonable as long as many different databases are involved in your data warehouse landscape. In this scenario you have to transport data from one place to another anyway, so it’s a legitimate way to do the transformation work in a separate specialized engine. Datastage is one of the most widely used ETL tools and demand for Datastage training is immense. Sky Infotech offers in-depth Datastage training in Noida and leads the pack of institutes eying best Datastage training institute in Noida tag.
Extract, Load, Transform (ELT) is a process where data is extracted from the source, then loaded into a staging table in the database, transforming it where it sits in the database and then loading it into the target database or data warehouse. When the data is extracted from the source into the staging tables, it is a raw copy i.e. you keep the column names the same as in the source database and you don’t convert data, calculate new data fields, etc. You may however filter unneeded rows and columns as you extract data so that you don’t waste resources on unneeded data.

ELT should be used, instead of ETL, in these various cases :

• The database engine is well adapted for that kind of processing
• There are big volumes of data
• The source database and the target database are the same
• So in short, when using ETL the transformations are processed by the ETL tools, while
in ELT the transformations are processed by the target datasource.

ELT has the benefit of minimizing the processing on the source since no transforming is being done. This can be extremely important if the source is a production system. The negative of this approach is it may take longer to get the data into the data warehouse as with the staging tables you have an extra step in the process, and you will need more disk space for the staging tables.