Amazon Timestream
Serverless time-series database for IoT, metrics, and operational telemetry
GreptimeDB
Unified observability database for metrics, logs, and traces
Cloud-native database built for real-time observability. Store and query metrics, logs, and traces in one system with sub-second performance at petabyte scale.
Overview
Real-Time Observability at Scale
GreptimeDB is a cloud-native database purpose-built for unified observability data collection and analysis. It consolidates metrics, logs, and traces into a single system, eliminating the complexity of managing separate data stores. Written in Rust with a distributed query engine and optimized columnar storage, it delivers sub-second query responses at petabyte scale.
Flexible Deployment, Powerful Querying
Designed for Kubernetes with compute/storage separation, GreptimeDB integrates natively with object storage (AWS S3, Azure Blob) and supports deployment from edge devices (including ARM/Android) to cloud environments. Developers can query data using SQL, PromQL, or streaming interfaces, and ingest through REST APIs, MySQL/PostgreSQL protocols, or popular observability protocols.
Production-Ready Architecture
Currently in beta with GA targeted for mid-2025, GreptimeDB is already deployed by early adopters in production environments. Its architecture leverages Apache Arrow for memory management, Parquet for storage, and DataFusion for query execution, ensuring exceptional cost efficiency and performance across hybrid and multi-cloud infrastructures.
Highlights
Unified storage for metrics, logs, and traces as timestamped wide events
Sub-second query performance at petabyte scale with distributed Rust engine
Cloud-native architecture with compute/storage separation and object storage support
Flexible deployment from edge devices to cloud with unified APIs
Pros
- Eliminates need for separate time-series, log, and trace databases
- High performance with Rust implementation and columnar storage optimization
- Multiple query interfaces: SQL, PromQL, and streaming support
- Deploy anywhere from ARM/Android edge devices to Kubernetes clusters
Considerations
- Currently in beta status with GA planned for mid-2025
- Requires Rust nightly toolchain and specific dependencies for building from source
- Newer project with smaller ecosystem compared to established observability tools
- Learning curve for teams unfamiliar with unified observability approaches
Managed products teams compare with
When teams consider GreptimeDB, 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
- Organizations consolidating observability infrastructure into a single platform
- Teams requiring real-time analytics on large-scale metrics, logs, and traces
- Edge-to-cloud deployments needing consistent APIs and efficient data sync
- Cloud-native environments leveraging Kubernetes and object storage
Not ideal when
- Projects requiring immediate production-grade stability guarantees before GA release
- Teams without capacity to evaluate beta software or contribute feedback
- Simple monitoring needs adequately served by existing single-purpose tools
- Environments with strict requirements for mature, long-established databases
How teams use it
Unified Observability Platform
Replace separate Prometheus, Elasticsearch, and tracing backends with one database, reducing operational complexity and infrastructure costs
Edge-to-Cloud IoT Monitoring
Deploy on ARM edge devices and sync telemetry to cloud clusters for real-time analytics across distributed sensor networks
High-Cardinality Metrics Analysis
Query petabyte-scale metrics with sub-second response times using SQL or PromQL for microservices observability
Multi-Cloud Observability
Leverage native object storage integration across AWS S3 and Azure Blob for cost-efficient, cross-cloud data access
Tech snapshot
Rust100%
Shell1%
TypeScript1%
Python1%
Dockerfile1%
Makefile1%
Frequently asked questions
What observability data types does GreptimeDB support?
GreptimeDB stores metrics, logs, and traces as unified timestamped events. You can query all three data types using SQL, PromQL, or streaming interfaces within a single database.
Is GreptimeDB ready for production use?
GreptimeDB is currently in beta and used by early adopters in production. GA (v1.0) is targeted for mid-2025. It's suitable for evaluation and pilot deployments, with the latest stable release recommended for production trials.
How does GreptimeDB achieve sub-second query performance?
Written in Rust with a distributed query engine, rich indexing, and optimized columnar storage based on Apache Arrow and Parquet, GreptimeDB delivers sub-second responses even at petabyte scale.
Can I deploy GreptimeDB on edge devices?
Yes, GreptimeDB supports flexible deployment from edge environments (including ARM and Android devices) to cloud infrastructure, with unified APIs and efficient data synchronization.
What protocols does GreptimeDB support for data ingestion?
GreptimeDB supports REST APIs, MySQL and PostgreSQL wire protocols, and popular observability ingestion protocols. SDKs are available for Go, Java, C++, Erlang, Rust, and JavaScript.
Project at a glance
Active- Stars
- 5,913
- Watchers
- 5,913
- Forks
- 449
LicenseApache-2.0
Repo age3 years old
Last commityesterday
Primary languageRust
Last synced yesterday