For one thing there is no single canonical way. Cassandra indexes (sstable indexes + bloom filters) are quite different than e.g. SQL server indexes (probably

For one thing there is no single canonical way. Cassandra indexes (sstable indexes + bloom filters) are quite different than e.g. SQL server indexes (probably some variant of B-trees). If I wasn't prepared I could probably say something about these schemes but database indexes are kind of specialized knowledge since they combine some attributed of in-memory data structures with the performance characteristics and limitations of storage. Let's turn the tables here, explain in detail how a B+ tree works and what's the difference between that and a "standard" B-tree? What do you know about sstables? LSMs?