Explore the security challenges and best practices for Apache Spark in this 43-minute conference talk from LASCON 2016. Dive into securing Spark through code and configuration, as well as integrations with commonly implemented technologies. Learn about the impact of various attacks against Spark and how to limit exposure. Examine ways to avoid common developer-induced issues, such as consuming untrusted serialized objects and misusing closures. Gain insights into data protection at-rest, in-memory, and over the network throughout its lifetime in a Spark ecosystem. Discover how to make better security decisions when implementing Spark in conjunction with distributed messaging systems like Kafka, high-throughput NoSQL databases like Cassandra, and resource management platforms like Mesos. By the end of this presentation, acquire the knowledge to use Spark more securely and effectively in big data analytics for stream and batch processing, machine learning, and predictive analytics.