ScyllaDB Manager 2.2: Repair Revisited

ScyllaDB Manager 2.2 introduces enhancements in its repair mechanisms, particularly optimizing for row-level repair in both the open-source and enterprise versions of ScyllaDB. The update includes a new parallel repair method that allows for faster repairs by optimizing the number of parallel repairs for a keyspace, beneficial for large clusters, and ensures each node participates in only one repair job at a time. The repair intensity feature has been refined, allowing adjustments based on node capabilities, and the ability to change repair speed in real-time without restarting tasks. Support for schema changes during repairs, small table optimizations, and a graceful stop feature have also been incorporated, ensuring smoother operation and management of repair processes. The update also features enhanced monitoring with detailed repair progress per table and keyspace, and a benchmark showing improved performance under various settings. These advancements make ScyllaDB Manager 2.2 a significant improvement for managing repairs efficiently, with resources available through the ScyllaDB Download Center and ScyllaDB University for further learning.