Why Elasticsearch’s 10,000 Hit Limit Slows Your Cluster and How to Fix It
Elasticsearch defaults to a total hit count of 10,000 after version 7.x, which many developers override with "track_total_hits": true to get exact numbers, but this seemingly harmless change can double CPU usage and increase query latency from 20 ms to 500 ms due to the underlying Block‑Max WAND algorithm and its interaction with aggregations, sorting, and scoring.
