Pipeline Aggregations which allow you to run aggregations on the results of other aggregations, including derivatives, moving averages, and Holt Winters predictive algorithms.
Better store compression to fit more data into your disks and file system cache.
Better heap usage with doc-values-by-default, reduced memory usage during merges, and roaring bitsets for filter caching.
Structured, readable, exceptions.
Optimised query execution order, auto-caching of filters, and parent-child rewritten for faster queries.
More>
Writes to transaction logs are atomic and durable by default.
A big cleanup to type mappings to make them safe, unambiguous, and>
Binding only to localhost by default.
Cluster state diffs for faster change propagation.