I was working in a project for business intelligence domain. We were using data warehouse as back-end. PL/SQL was used for implementation of ETL. Some new architect suggested that we should either move to a tool like informatica or use Hadoop. The only problem was we didn't know if hadoop will work for as an alternative of Data Warehouse. So was curious when we say Hadoop Real world solutions. Does it cover solutions like this? Where we have a legacy data warehouse implementation and we are facing performance issues.
With regards to your problem, Hadoop Real world solutions is designed to get people up and rolling with Hadoop once they've decided it's a good fit for their system. There is no one particular recipe that covers a full transition from a legacy warehouse to using Hadoop, but many of the recipes independently help with a critical piece. For example, we cover using Sqoop to migrate data from a RDBMS (like Oracle) into Hadoop HDFS. That would be a good place to start.
Hive/Pig are excellent data warehousing/processing tools designed to run over Hadoop, and we cover how to use them in very solid detail. You can follow the sample analytics to get an idea of how you would port your PL/SQL implementations to a language like Pig Latin or HiveQL.
So to answer your question, if you've decided Hadoop is the right platform moving forward, many of the Real-World solutions recipes can be combined to help with the overall data migration and ETL transition.