Fuss is regularly made about inefficient schema evolution in RDBMS. Just throwing data as textfiles into Hadoop is not really the solution. With Hadoop, you get many choices about file formats. Avro is a choice that allows schema evolution as described by Gwen Shapira in “The problem of managing schemas“.
- “DZone Best of the Year: NoSQL Zone Edition” by G. Ryan Spain (DZone blog)
- “Top Ten Popular Hadoop Blog Posts of 2014” by Jules S. Damji (Hortonworks blog)
- “Top 10 Hadoop Blogs of 2014” by Karen Whipple (MapR blog)
- “The Top 10 Posts of 2014 from the Cloudera Engineering Blog” by Justin Kestelyn (Cloudera Blog)