Beyond filter rewrite: How doc value skip indexes accelerate aggregations in OpenSearch

This article details how the authors overcame limitations in previous date histogram aggregation optimizations in Elasticsearch. Previously, these optimizations (filter rewrite & multi-range traversal) required filtering and aggregation on the same field and couldn't handle complex subaggregations. They’ve now leveraged Lucene 10’s doc value skip indexes – a hierarchical summary of numeric data – to achieve up to 28x faster aggregations even when filter and aggregation fields are different, and to support complex subaggregations by enabling bulk counting or skipping of document intervals. This new approach operates directly on the aggregation field’s doc values, decoupling it from the filter and significantly improving performance for common analytical queries.