Algolia
Hosted search-as-a-service platform delivering real-time, full-text search for apps and websites
Quickwit
Fast, cloud-native search engine for logs and traces
Quickwit delivers sub‑second, full‑text search and analytics on cloud storage for log management, distributed tracing, and upcoming metrics, with Elasticsearch‑compatible APIs and native Grafana integration.
Overview
Who should use Quickwit?
Quickwit is built for engineers and ops teams that need high‑performance search over observability data without the operational overhead of traditional search clusters. It fits organizations moving to cloud‑native stacks, those consolidating logs and traces, and teams looking to cut costs while keeping Elasticsearch‑compatible query capabilities.
Core capabilities
The engine indexes data directly on object stores such as Amazon S3, Azure Blob, or Google Cloud Storage, delivering sub‑second query latency. It supports full‑text search, aggregations, and schemaless or strict schema indexing. Native integrations include Jaeger for tracing, OpenTelemetry for logs and traces, and a Grafana data source for visual analytics. The Elasticsearch‑compatible REST API lets you reuse existing clients and dashboards.
Deployment and scalability
Quickwit runs stateless searchers and indexers, decoupling compute from storage. It is Kubernetes‑ready with an official Helm chart, and can ingest from Kafka, Kinesis, or Pulsar for multi‑tenant, high‑availability setups. Retention policies and GDPR‑oriented delete tasks are built‑in, making long‑term data management straightforward.
Highlights
Sub‑second search on cloud storage (S3, Azure, GCS)
Elasticsearch‑compatible API for seamless migration
Jaeger and OpenTelemetry native ingestion
Stateless, decoupled compute and storage with Kubernetes‑ready helm chart
Pros
- High query performance on object storage
- Cost‑effective compared to traditional search stacks
- Supports existing Elasticsearch clients and tools
- Built‑in observability integrations (Jaeger, OTEL, Grafana)
Considerations
- Metrics support is still on the roadmap
- HA indexing requires a Kafka source
- Limited to cloud storage backends
- Feature set may lag behind mature Elasticsearch plugins
Managed products teams compare with
When teams consider Quickwit, these hosted platforms usually appear on the same shortlist.
Looking for a hosted option? These are the services engineering teams benchmark against before choosing open source.
Fit guide
Great for
- Teams needing fast log search without managing Elasticsearch clusters
- Organizations adopting cloud‑native observability pipelines
- Developers wanting Elasticsearch‑compatible API on S3
- Ops teams requiring multi‑tenant, retention‑policy aware storage
Not ideal when
- Workloads requiring on‑premise storage only
- Environments needing immediate metrics ingestion
- Deployments without Kafka for HA indexing
- Use cases demanding extensive Elasticsearch plugin ecosystem
How teams use it
Centralized log analysis for microservices
Ingest logs from Kubernetes via Fluent Bit, query across all services in sub‑second time, and reduce storage costs.
Distributed tracing with Jaeger
Collect trace spans, run ad‑hoc trace queries through Grafana, and accelerate root‑cause analysis.
Migration from Elasticsearch
Switch ingest pipelines to Quickwit using the compatible API, retain existing dashboards, and cut infrastructure spend.
Retention‑policy driven archiving
Define per‑index retention, automatically purge old data to meet compliance and GDPR requirements.
Tech snapshot
Rust87%
HTML10%
TypeScript1%
HCL1%
Shell1%
JavaScript1%
Frequently asked questions
How can I switch from Elasticsearch or OpenSearch to Quickwit?
Quickwit implements a large subset of the Elasticsearch/OpenSearch API, including ingest endpoints and popular query DSL features, allowing you to migrate log shippers and clients with minimal changes.
How is Quickwit different from traditional search engines like Elasticsearch or Solr?
Quickwit is architected to search directly on cloud object storage, optimizing I/O paths and using stateless indexers and searchers, which yields sub‑second latency without managing heavy storage nodes.
How does Quickwit compare to Elastic in terms of cost?
The project estimates up to 10× lower total cost of ownership by leveraging cheap cloud storage instead of dedicated search hardware.
What license does Quickwit use?
Quickwit is released under the Apache License, Version 2.0.
Is it possible to set up Quickwit for a High Availability (HA) configuration?
HA is available for search operations out of the box; HA indexing requires a Kafka source.
Project at a glance
Active- Stars
- 10,788
- Watchers
- 10,788
- Forks
- 508
LicenseApache-2.0
Repo age4 years old
Last commit5 days ago
Primary languageRust
Last synced yesterday