Distributed database systems are extensively utilized to provide cloud services for online map platforms, offering consistency, disaster recovery, and high performance, whereas traditional systems relying on singly-homed architecture face challenges in scaling for large-scale services. In this paper, we propose the architectural design of OceanBase (OB), a distributed database system that “unitizes” services and operations into individual machines. The unitization approach migrates from a singly-homed to a multi-homed design across multiple regions. By leveraging this feature, OceanBase ensures data replication and seamless service handover when a machine goes offline. However, communication overhead between regions can sometimes be burdensome. To address this issue, OceanBase unitizes read and write operations, and employs a hybrid centralization and unitization approach that is dynamically optimized for Online Transaction Processing (OLTP) and Online Analytical Processing (OLAP). To validate our design, we deploy OceanBase on AMap, an online map application platform supporting large-scale distributed services. Through a series of experiments, we demonstrate that OceanBase exhibits enhanced disaster tolerance capabilities and achieves improved performance for both write-intensive and read-intensive benchmarks.

Distributed databases have been widely researched and developed in recent years due to their scalability, availability, and consistency guarantees. The write-ahead logging (WAL) system is one of the most vital components in a database. It is still a non-trivial problem to design a replicated logging system as the foundation of a distributed database with the power of ACID transactions. This paper proposes PALF, a Paxos-backed Append-only Log File System, to address these challenges. The basic idea behind PALF is to co-design the logging system with the entire database for supporting database-specific functions and to abstract the functions as PALF primitives to power other distributed systems. Many database functions, including transaction processing, database restore, and physical standby databases, have been built based on PALF primitives. Evaluation shows that PALF greatly outperforms well-known implementations of consensus protocols and is fully competent for distributed database workloads. PALF has been deployed as a component of the OceanBase 4.0 database and has been made open-source along with it.

The problem of deadlock detection and resolution in database systems has been studied for decades. While it has long been a mature feature of classical centralized database systems for many years, its use in distributed database systems remains in its infancy. Don P. Mitchell and Michael J. Merritt (M&M) proposed a simple and fully distributed deadlock detection and resolution algorithm, but its assumption that each process waits on only one resource at a time prevents it from being generally applicable. Inspired by this algorithm, we design and implement LCL (Lock Chain Length), an elegant and generally applicable algorithm for resource deadlock detection and resolution in distributed environments without a restriction of the above kind. Our extensive emulation experiments show that the proposed approach LCL significantly outperforms the state-of-the-art competitor M&M. In addition, it has been applied to the OceanBase distributed relational database system, and our extensive experiments in OceanBase illustrate that LCL is also more efficient than M&M in deadlock detection and resolution.

Countrywide and worldwide business, like gaming and social networks, drives the popularity of inter-data-center transactions. To support inter-data-center transaction processing and data center fault tolerance simultaneously, existing protocols suffer from significant performance degradation due to high-latency and unstable networks. In this paper, we propose RedT, a novel distributed transaction processing protocol that works in heterogeneous networks. In detail, nodes within a data center are inter-connected via the RDMA-capable network and nodes across data centers are inter-connected via TCP/IP networks. RedT extends two-phase commit (2PC) by decomposing transactions into sub-transactions in terms of the data center granularity, and proposing a pre-write-log mechanism that is able to reduce the number of inter-data-center round-trips from a maximal of 6 to 2. Extensive evaluation against state-of-the-art protocols shows that RedT can achieve up to 1.57× higher throughputs and 0.56× lower latency.

We have designed and developed OceanBase, a distributed relational database system from the very basics for a decade. Being a scale-out multi-tenant system, OceanBase is cross-region fault tolerant, which is based on the shared-nothing architecture. Besides sharing many similar goals with alternative distributed DBMS, such as horizontal scalability, fault-tolerance, etc., our design has been driven by the demands of typical RDBMS compatibility as well as both on-premise and off-premise deployments. OceanBase has fulfilled its design goal. It implements the salient features of certain mainstream classical RDBMS, and most applications on them can run on OceanBase, with or without a few minor modifications. Tens of thousands of OceanBase servers have been deployed in Alipay.com as well as many other commercial organizations. It has also successfully passed the TPC-C benchmark test and seized the first place with more than 707 million tpmC. This paper presents the goals, design criteria, infrastructure, and key components of OceanBase including its engines for storage and transaction processing. Further, it details how OceanBase achieves the above leading TPC-C benchmark in a distributed cluster with more than 1,500 servers from 3 zones. It also describes lessons what we have learnt in building OceanBase for more than a decade.