Shark is a research data analysis system built on a novel coarse-grained distributed shared-memory abstraction. Shark marries query processing with deep data analysis, providing a unified system for easy data manipulation using SQL and pushing sophisticated analysis closer to data. It scales to thousands of nodes in a fault-tolerant manner. Shark can answer queries 40X faster than Apache Hive and run machine learning programs 25X faster than MapReduce programs in Apache Hadoop on large datasets. This is a complete overview of the development of Shark, including design decisions, performance details, and comparison with existing data warehousing solutions. It demonstrates some of Shark's distinguishing features including its in-memory columnar caching and its unified machine learning interface.