Restricted access

June 25, 2008

ETL - migration puzzle

Filed under: Data Migration, ETL — Alena Semeshko @ 6:32 am

John Morris’s latest post, where he highlights the difference between data migration and ETL is, in fact, very much to the point. ETL is a much larger concept that needs more effort and time than data migration. Data migration, if properly defined, is a one-way long-term trip of data from one system to another. While ETL not only can be a repeated procedure, but is also much larger in scope, as it includes all sorts of data transformations (from data cleansing and data quality checking to certain elements of business intelligence).

So, if you look closely, the two are considerably different, not even parts of the same process.

1 Comment »

  1. Alena,

    You are possibly right about the size of concept; ETL is deployed for many data problems. However, I think Johny’s point is related to the fact that ETL is not a good standard for data migration problems when the size or complexity of the data or business impact is large. ETL and especially ETL deployed to write to data warehouses is a much smaller problem space than data migration.

    A significant difference is that data migration has a major impact on business process (especially during the transformation period). ETL generally does not consider business process and or business process change and hence is an inappropriate method/tool-set for all but the smallest data migration problems.


    Comment by Paul Hollingsworth — June 26, 2008 @ 5:08 am

RSS feed for comments on this post. TrackBack URL

Leave a comment