DuyHai Doan takes us inside of Apache Kudu, a data store designed to support fast access for analytics in the Hadoop ecosystem. We compare Kudu’s architecture with Apache Cassandra and discuss why effective design patterns for distributed systems show up again and again.