Quickwit
activeby Quickwit (Datadog) · since 2021
Cloud-native search engine on object storage for logs and traces, built on Tantivy.
Overview
Quickwit is a Rust search engine designed for append-only observability data: logs, traces and events. It separates compute from storage, keeping indexes on S3-compatible object storage, so stateless searchers can scale out and retention costs a fraction of an Elasticsearch hot tier. It offers an Elasticsearch-compatible query API, native Jaeger and OpenTelemetry endpoints, and Kafka/Kinesis/Pulsar ingestion. Datadog acquired Quickwit in early 2025 and relicensed it from AGPL to Apache-2.0; development continues in the open with the 0.9 line.
Strengths & trade-offs
Strengths
- +Indexes live on object storage, so long retention is cheap
- +Stateless searchers scale independently of ingest
- +Native OpenTelemetry and Jaeger endpoints for traces and logs
- +Elasticsearch-compatible search API works with Grafana and existing tools
- +Apache-2.0 after the Datadog acquisition
Trade-offs
- −Optimised for append-only data; updates and deletes are limited
- −No vector search
- −Object-storage latency makes it slower than RAM-heavy engines for interactive app search
- −Future roadmap depends on Datadog's priorities
Great for
- +Log search and retention at large scale
- +Distributed tracing backend
- +Security and audit event search
- +Cheap replacement for Elasticsearch hot-warm logging clusters
Think twice for
- −E-commerce or app search with frequent updates
- −Vector/RAG retrieval
- −Transactional data
Capabilities
Native or official first-party support · 7 of 25 tracked
Scale & consistency
- Comfortable data size
- Massive · Multi-PB
- Horizontal scaling
- Native
- Multi-region
- No
- Transactions
- None
- Consistency
- Eventual
- Schema
- Flexible
- Architecture
- Distributed, Shared storage
- Query languages
- Quickwit query language, Elasticsearch Query DSL (subset)
Stateless indexers and searchers over object storage; Binance runs a 100 PB log service ingesting about 1.6 PB per day.
Deployment
Ecosystem
Alternatives to Quickwit
Lucene-based distributed search and analytics engine for text, vectors, logs and metrics.
quickwit vs elasticsearch →Apache-licensed, foundation-governed fork of Elasticsearch for search, vectors and observability.
quickwit vs opensearch →Label-indexed log aggregation: Prometheus-style streams, chunks on object storage, LogQL queries.
quickwit vs grafana-loki →Single-binary Rust observability platform storing logs, metrics and traces as Parquet on object storage.
quickwit vs openobserve →Log database from the VictoriaMetrics team: schemaless ingestion, LogsQL and low RAM and disk use.
quickwit vs victorialogs →Open-source columnar OLAP database known for very fast aggregations over billions of rows.
quickwit vs clickhouse →Quickwit FAQ
What is Quickwit?
Quickwit is a search engine database by Quickwit (Datadog), first released in 2021. Quickwit is a Rust search engine designed for append-only observability data: logs, traces and events. It separates compute from storage, keeping indexes on S3-compatible object storage, so stateless searchers can scale out and retention costs a fraction of an Elasticsearch hot tier. It offers an Elasticsearch-compatible query API, native Jaeger and OpenTelemetry endpoints, and Kafka/Kinesis/Pulsar ingestion. Datadog acquired Quickwit in early 2025 and relicensed it from AGPL to Apache-2.0; development continues in the open with the 0.9 line.
Is Quickwit open source?
Yes. Quickwit is open source. License: Apache License 2.0.
Does Quickwit support ACID transactions?
No. Quickwit does not provide ACID transactions; its consistency model is eventual.
Does Quickwit scale horizontally?
Quickwit scales horizontally out of the box, with built-in sharding or partitioning. It is comfortable at massive scale (Multi-PB).
How can I run Quickwit?
You can self-host it.
What is Quickwit best for?
Quickwit is a strong fit for log search and retention at large scale; distributed tracing backend; and security and audit event search. Think twice for E-commerce or app search with frequent updates.
What are the best alternatives to Quickwit?
Common alternatives to Quickwit include Elasticsearch, OpenSearch, Grafana Loki, OpenObserve and VictoriaLogs. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
How popular is Quickwit?
Quickwit ranks #275 of 844 databases on the WhichDB Score (11.9/100), and #19 among search engine databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.