apache-orc

Smallest, fastest columnar storage for Hadoop workloads

Apache ORC is a self-describing type-aware columnar file format designed for Hadoop workloads. It is optimized for large streaming reads, but with integrated support for finding required rows quickly.