Do you still need Elasticsearch for log analytics? ClickHouse says no.

ClickHouse now includes redesigned full-text search powered by inverted indexes, making it a viable alternative to Elasticsearch for log analytics. A detailed benchmark compares both systems on a realistic OpenTelemetry log workload at 1B, 10B, and 50B rows. Results show ClickHouse is 2–6× faster on cold queries and 1.7–2.6× faster on hot queries, while storing the same data in roughly one-fifth the space (4.95× less). The performance advantage grows with workloads that combine full-text search with aggregations — typical in log analysis. Ingestion speed also favors ClickHouse dramatically: loading 50B rows took under 4 hours vs. ~5 days for Elasticsearch. The benchmark is fully reproducible with source data, schemas, and queries published on GitHub.