Skip to content

GreptimeDB Open Source

One open-source database for metrics, logs, traces, and wide events. Deploy as standalone or cluster. PromQL + SQL, object storage first, up to 50x cost reduction.

curl -L https://github.com/GreptimeTeam/greptimedb/raw/main/scripts/install.sh | sh

Integrate GreptimeDB with your preferred technology stack

Installation Guide

Learn more
ingest-data/for-observability/opentelemetryingest-data/for-observability/fluent-bitingest-data/for-observability/elasticsearchingest-data/for-observability/lokiingest-data/for-observability/kafkaingest-data/for-iot/grpc-sdks/goingest-data/for-iot/grpc-sdks/javaingest-data/for-iot/emqxintegrations/prometheusintegrations/grafanaprotocols/mysqlprotocols/influxdb-line-protocolprotocols/postgresql

Why GreptimeDB

Unified Observability Engine

Unified Observability Engine

Metrics, logs, traces, and wide events in one database. OpenTelemetry-native ingestion, SQL + PromQL query, Flow engine for continuous aggregation and metric derivation. One system replaces Prometheus + Loki + Elasticsearch.

Object Storage First — Up to 50x Cost Reduction

Object Storage First — Up to 50x Cost Reduction

Compute-storage disaggregation architecture with object storage (S3, Azure Blob, GCS) as primary. Scale storage and compute independently. Pay for what you use — no over-provisioning.

Drop-In Compatibility

Drop-In Compatibility

PromQL-compatible for existing Prometheus dashboards and alerts. Loki Push API and Elasticsearch bulk API for log ingestion. Jaeger Query API for trace visualization. Grafana datasource, MySQL/PostgreSQL wire protocol. Migrate without rewriting.

High-Performance Columnar Engine

High-Performance Columnar Engine

Built with Rust on Apache Arrow and DataFusion. Rich indexing (inverted, fulltext, skipping) enables sub-second queries on petabyte-scale datasets. Handles high-dimensional, high-cardinality observability events.

Stateless Scale-Out

Stateless Scale-Out

Designed for Kubernetes. Stateless compute nodes scale horizontally — add nodes to scale, remove without data loss. No Thanos-style sidecar complexity.

Edge to Cloud

Edge to Cloud

Deploy on ARM-based edge devices or cloud clusters with identical APIs. Bandwidth-efficient sync between edge and cloud. One binary, from a Raspberry Pi to a 100-node cluster.

When to consider upgrading

Scaling beyond development and testing

Enterprise clustering and professional support

Learn More

Need to eliminate operational overhead

Fully-managed cloud service with auto-optimization

Contact Us

Building edge-cloud solutions

Specialized automotive and IoT integration

Learn More

Stay in the loop

Join our community