YDB is a distributed SQL database that combines high availability and scalability with strong consistency and ACID transactions.
YDB addresses the need for scalable interactive web services that require both transactional consistency and horizontal scaling. It achieves this through a disaggregated architecture separating storage and compute layers, allowing each to scale independently. The system maintains strict consistency and ACID guarantees across multiple nodes and tables, supporting both row-oriented and column-oriented table formats alongside persistent queues for data movement. It survives complete outages of individual datacenters while remaining available for reads and writes, with automatic recovery restoring data redundancy after hardware failures.
YDB suits teams building OLTP-like workloads that demand both scalability and transactional guarantees, particularly those operating at scale across multiple availability zones. It offers a rich SQL dialect called YQL for data manipulation and schema definition, with PostgreSQL-compatible mode for table operations and Kafka-compatible mode for topics. Deployment options include Ansible, Kubernetes, or manual configuration. The system is designed for production environments handling millions of distributed transactions per second across thousands of nodes storing petabytes of data.
The project maintains active development across its distributed systems and database components, with ongoing work on core transaction processing and fault tolerance mechanisms. The codebase shows consistent refinement of storage layer optimization and cluster coordination logic. Community engagement remains steady through multiple channels including Discord and Telegram, with documentation and examples receiving regular updates to reflect the system's capabilities.
