OpenObserve

OpenObserve (O2) is a cloud-native, open source observability platform that unifies logs, metrics, traces, frontend monitoring, and session replay into a single solution. Written in Rust and built on Apache Parquet with the DataFusion query engine, it delivers petabyte-scale performance at dramatically lower cost than traditional tools like Elasticsearch. The platform is licensed under AGPL-3.0 and supports both cloud-hosted and self-hosted deployments.

• Unified Observability — Ingest and query logs, metrics, traces, RUM, session replay, and error tracking from one platform.
• 140x Lower Storage Cost — High compression (~40x) via columnar Apache Parquet storage benchmarked against Elasticsearch.
• Blazing Performance — Rust-native engine with DataFusion queries; internal benchmarks return 1 petabyte of data in 2 seconds.
• Petabyte Scalability — Stateless node architecture enables horizontal scaling without data complexity.
• Bring Your Own Bucket — Supports local disk, S3, MinIO, GCS, and Azure Blob Storage for long-term data retention.
• Data Pipelines — Configure ingestion-time data processing workflows including sensitive data redaction and multi-destination routing.
• Alerts & Dashboards — Set up continuous monitoring alerts and build rich visualizations for real-time and historical data.
• OpenTelemetry Compatible — Embraces open standards with vendor-neutral API interfaces and OTel-native ingestion.
• Role-Based Access Control — Manage user identities and fine-grained resource access via built-in IAM.
• O2 AI Assistant — Built-in AI assistant for accelerating observability workflows and query analysis.
• Self-Hosted or Cloud — Get started in under 2 minutes with Docker locally, or use the managed cloud offering with a 14-day free trial.