The system architecture section provides an in-depth overview of the overall architecture, storage architecture, database security, data reliability, and high availability of OceanBase Database. For more information, see OceanBase concepts.
OceanBase Database uses a shared-nothing distributed architecture, where nodes are equal and each node has its own SQL engine, transaction engine, and storage engine. A database cluster consists of multiple zones, each containing multiple OBServer nodes that handle SQL queries, transactions, and storage. OceanBase Database supports multi-tenancy, with each tenant being an independent database instance with isolated resources.
Storage architecture
OceanBase Database uses an LSM-tree-based storage engine, dividing data into static baseline data (SSTables) and dynamic incremental data (MemTables). Data is first written to a MemTable and then flushed to disk as an SSTable when it reaches a certain size. Queries simultaneously access both the MemTable and SSTable, with results merged. The database optimizes query performance using block cache and row cache and employs aggressive compression algorithms to reduce storage costs.
Database security
OceanBase Database provides a comprehensive security system, including identity authentication, access control, data encryption, monitoring alerts, and security auditing. Supports full-link data encryption and transparent storage encryption to ensure data security during transmission and storage.
Data reliability and high availability
OceanBase Database ensures data consistency and high availability through multi-replica mechanisms and the Paxos protocol. Supports server-level, IDC-level, and region-level disaster recovery, achieving high availability standards with RPO=0 and RTO<8 seconds.
Database objects
OceanBase Database stores database object data in SSTables, supporting automatic tablespace management. Users do not need to explicitly create or manage tablespaces. These features collectively form the core system architecture of OceanBase Database, showcasing its technical advantages in distributed architecture, storage optimization, security, and high availability.