DB2 for z/OS is no longer about the traditional relational structured data only. The evolution brought us not only binary objects, or XML and JSON, but also other non-structured data living in a distributed file system such as HDFS.
Apache Spark has become a hot topic in the world of Analytics. It is seen as a better performing alternative to Map/Reduce , and as a means to do analytics combining data across a wide set of data sources. This video is a excellent introduction to what Spark is, and how it works with DB2 for z/OS.
The Logical Data Warehouse is a data management architecture for analytics. It is a collection of data repositories and associated data services for management and access to data. IBM Fluid Query is a new software component that extends the concept of a federated query to Big Data.
The modern enterprise data warehouse needs to be a heterogeneous mix of integrated software to handle a wide variety of data sources both structured, semi-structured and unstructured. The combination of DB2 with Hadoop is one approach that meets the needs for this diverse set of data.
Here are some links from IDUG's big data primer article for quick reference