Skip to content
GitHub
6.1k
v1.0.0-rc.2

The Single Database for Observability

The only database that speaks your entire observability stack, metrics, logs, traces and wide events, unified through OpenTelemetry, queryable with SQL and PromQL, built to scale on object storage.

Streaming
Storage

OBSERVABILITY 1.0

Three systems, three problems

  • Separate ingestion pipelines per signal
  • Pre-aggregation discards detail before incidents hit
  • No cross-signal correlation in a single query
  • Scale = more components, more local disk sprawl
  • Three dashboards, three alert configs
VS
OBSERVABILITY 2.0 →

One engine, Wide Events

  • All signals are timestamped events at their core
  • Store raw events, derive any metric on demand
  • JOIN metrics + logs + traces in one SQL query
  • Stateless compute on object storage, up to 50x cheaper
  • One datasource, one operational surface

The Unified Database Powers Your Observability Stack

Unify metrics, logs, traces, and wide events through standard protocols and a high-speed ingestion pipeline. Built-in streaming, materialized views, and a distributed query engine, all on object storage. An Observability 2.0 database from the ground up.

Data sources
Metrics
Wide Events
Traces
Logs
Standard
Open Protocols
OpenTelemetry
Prometheus Remote Write
Ultra high-speed
Ingestion
In-Database Computing
Preprocessing
Pipeline
Streaming &
Materialized View
Alerting/Trigger Rules
GreptimeDB
Compacted and Compressed Low-cost Scalable Data Storage
Object Storage
Applications
APM
Dashboard
O11y Data Analytics
AI & Machine Learning
Alerting
Standard
Protocols & Drivers
SQL
PromQL
Fast and Flexible
Query options

GreptimeDB is Trusted By

Migrate Loki to GreptimeDB
Oceanbase
Staff Engineer
Learn more
Migrating from Loki to GreptimeDB enables high-performance querying of massive log data at scale, offers multi-cloud deployment flexibility, and significantly simplifies application and deployment architecture.
Query Performance
10xincrease
TCO Reduction
30%decrease
Timeout Elimination
With sub-second response

Why the architecture works

Every design choice follows from one insight — observability signals are all timestamped events.

Wide Events demand columnar + object storage

Wide Events demand columnar + object storage

  • High-dimensional events with hundreds of fields per row
  • Columnar engine compresses and queries sparse columns efficiently
  • Object storage as primary, compute scales independently
Traffic-driven data volume demands elastic infrastructure

Traffic-driven data volume demands elastic infrastructure

  • Wide events grow with your traffic, not with your config
  • Compute-storage disaggregation — add compute for spikes, storage grows on S3
  • Sub-second data visibility, even during ingestion bursts
Unified events demand unified query + derivation

Unified events demand unified query + derivation

  • SQL + PromQL for both routine dashboards and ad-hoc exploration
  • JOIN metrics, logs, and traces in one query
  • Derive metrics from raw events via Flow engine
Built on Rust, Arrow & DataFusion

Built on Rust, Arrow & DataFusion

  • Memory-safe, high-performance Rust foundation
  • Apache Arrow and DataFusion for vectorized query execution
  • Open data formats — no vendor lock-in

Plan That Fits Your Needs

Enterprise User?

Contact us
Open source

Open source

One database for metrics, logs, traces, and wide events. SQL + PromQL, object storage first, Apache 2.0 licensed.

  • Production-ready standalone or cluster deployment

  • Full observability stack with Flow engine for metric derivation

  • Community-driven, open governance

Enterprise

Enterprise

Enterprise-grade observability data platform for business-critical applications with enhanced security, high availability, and professional support.

  • Mission-critical systems monitoring and analytics

  • Higher performance and data governance

  • Industries requiring high security and compliance (finance, healthcare etc.)

  • Organizations needing automated operations and intelligent resource scheduling

  • Teams that prefer self-hosting with enterprise support