EXTRACT, TRANSFORM AND LOAD (ETL)

ETL is the first tier of the required information infrastructure. The ETL is the process that routinely extracts and loads the relational data warehouse or datamart(s). 

• Data and business rules held in back-end systems may need to be applied as data is loaded to ensure downstream data consistency.
• This is generally accomplished via SQL Server Integration Services (SSIS) packages that execute processes to collect data from various organizational data sources during periods of minimal activity.
• These packages are physically stored on a relational database server and operate in a ‘pull mode’. By executing the packages on the database server, ComFrame BI architects allocate the majority of the ETL workload onto the data warehouse server and not on the company’s production servers.
• A separate SSIS server can be optionally allocated if high volumes of data are continuously being processed in the ETL.