Link to a developerWorks blog post which discusses how to represent complex data types in Big SQL tables through the use of SerDes.
Hadoop supports a diverse variety of data sources from flat text files to different databases. Dealing with semi-structured and unstructured data brings in the complexity of handling many custom data formats which are not easily handled by the traditional RDBMSs. IBM InfoSphere BigInsights provides the most familiar SQL-like interface with its Big SQL technology to work with Hadoop data within the BigInsights environment. With its support for complex data types and custom Serializers/Deserializers (SerDe) like Hive, Big SQL is quite efficient in dealing with custom file formats. We shall now see how we can represent complex data in Big SQL tables using a SerDe interface.
Author: Roopa M Sivakumar