HoraeDB Joins Apache Incubator: Design Goals, Architecture, and Core Features of the Cloud‑Native Time‑Series Database
HoraeDB, the next‑generation cloud‑native time‑series database originally from Ant Group, has been accepted into the Apache Incubator, and this article outlines its design motivations, distributed architecture, key technical components, and core capabilities such as high performance, low cost, SQL‑based analytics, storage‑compute separation, high availability, and open‑source ecosystem compatibility.
Ant Group's time‑series database CeresDB, rebranded as HoraeDB, was officially donated to the Apache Software Foundation on December 11, 2023 and unanimously approved to join the Apache Incubator.
After joining, HoraeDB commits to the Apache Way of openness and collaboration, encouraging community contributions and aiming to build a fair, diverse, and inclusive project.
Design Goals
The team, with five years of experience in time‑series data, identified challenges such as efficient handling of high‑cardinality tag combinations, modern distributed solutions, and cloud‑native compute‑storage separation. Consequently, HoraeDB is defined as a next‑generation cloud‑native time‑series database that supports both traditional time‑series workloads and analytic workloads across various deployment scales.
Architecture Overview
HoraeDB adopts a storage‑compute separation architecture. The system consists of:
HoraeMeta Cluster – metadata center for overall scheduling.
HoraeDB instances – responsible for time‑series data organization and storage.
WAL Service (external) – stores real‑time write‑ahead logs.
Object Storage (external) – persists SST files generated from memtables.
This design enables elastic scaling of compute and storage, high availability, and load balancing.
Core Features
High performance, low cost: a custom Rust engine with a 10:1 compression ratio.
SQL‑based OLAP analytics with push‑down of predicates and aggregations.
Storage‑compute separation supporting multiple storage media and tiered (hot/cold) data.
Horizontal scalability for data shards, compute, and storage, plus cross‑region disaster recovery meeting financial‑grade standards.
Rich data model: Metric + Tags + MultiFields.
RESTful API and multi‑language SDKs.
Compatibility with Prometheus, InfluxDB, and OpenTSDB protocols.
Community and Acknowledgments
The project thanks its contributors, champions, and mentors, especially the Ant Group database team, for their support.
Join the Community
Developers and users interested in HoraeDB are invited to contribute via the GitHub repository (https://github.com/apache/incubator-horaedb.git), the official website (https://horaedb.apache.org), the mailing list ([email protected]), and the WeChat channels ApacheHoraeDB.
AntTech
Technology is the core driver of Ant's future creation.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.