Apache Big Data Europe 2016

As big data continues to get bigger, deploying flexible and robust security is more important than ever. In this talk, we'll discuss about Apache Sentry, a central service for policy management and its various pluggable authorization engines which integrates with many Hadoop components. And we will dive deep into how its latest design allows for fault tolerance, high availability and scalability.



Unlike traditional database systems, authorization in Hadoop eco system is a tricky problem due to the fact that there are multiple doors to the same data. Sentry provides a great deal of usability by letting users define policies once and it replicates the state as necessary. With this comes additional challenges of designing a distributed service which manages consistent state. This talk will touch upon core design choices which lie as building blocks to any robust distributed system.