Overview:
Quickwit is a cloud-native search engine designed for observability use cases, specifically log management and distributed tracing. It is positioned as an open-source alternative to Datadog, Elasticsearch, Loki, and Tempo. The project focuses on enabling sub-second full-text search and aggregation queries directly on cloud storage, using a decoupled compute and storage architecture. It is built for teams and developers who need to manage and analyze large volumes of logs and traces at scale without relying on traditional, storage-bound search infrastructure.
Core Features:
Full-text search and aggregation queries: Supports search and analytics on indexed data with sub-second latency on cloud storage.
Elasticsearch-compatible API: Provides an API compatible with Elasticsearch and OpenSearch clients for search, ingestion, and query DSL.
Jaeger-native and OTEL-native: Built-in support for Jaeger distributed tracing and OpenTelemetry-native ingestion for logs and traces.
Schemaless or strict schema indexing: Allows indexing data with flexible schemas or enforcing a strict schema.
Multi-tenancy with retention policies: Supports indexing across many indexes with partitioning and configurable retention policies.
Kubernetes ready: Includes a Helm chart for deployment on Kubernetes clusters.
Use Cases:
Log management at scale: Teams can ingest and search large volumes of log data stored on Amazon S3, Azure Blob Storage, or Google Cloud Storage.
Distributed tracing analysis: Developers can use Quickwit as a backend for Jaeger or Grafana to analyze traces from OTEL-native sources.
GDPR compliance workflows: Administrators can use delete tasks to remove specific indexed data for regulatory requirements.
Migrating from Elasticsearch: Organizations using Elasticsearch or OpenSearch can transition log shippers (e.g., Vector, Fluent Bit) to Quickwit using its compatible ingest and search API.
Why It Matters:
Quickwit offers a search engine built from the ground up for cloud storage, which allows it to operate with stateless indexers and searchers. For teams managing observability data, this architecture can reduce storage costs compared to traditional search engines like Elasticsearch. It provides an open-source alternative to proprietary observability platforms without requiring users to manage local storage or stateful clusters. Its compatibility with Elasticsearch, Jaeger, and OpenTelemetry makes it a practical drop-in for existing observability pipelines.
