2019/03/01 by Yi Lu, Xiangyao Yu, Lu, Yi +3
Computer Science · #Advanced Data Storage Technologies #Artificial intelligence #Asynchronous communication #Backup #Computer network #Computer science #Concurrency #Concurrency control #Consistency (knowledge bases) #Database #Database transaction #Databases (cs.DB) #Distributed and Parallel Computing Systems #Distributed computing #Distributed systems and fault tolerance #Distributed transaction #FOS: Computer and information sciences #Operating system #Replica #Replication (statistics) #Serializability #Synchronization (alternating current) #Throughput #Timestamp #Transaction processing #Wireless #cs.DB
paper · pdf · doi:10.48550/arxiv.1903.00424
published in arXiv (Cornell University) (Cornell University)
arxiv created 2019/03/01 · openalex publication_date 2019/03/01 · arxiv updated 2019/03/04 · openalex created_date 2019/03/11 · openalex updated_date 2026/07/28
Data replication is crucial in modern distributed systems as a means to provide high availability. Many techniques have been proposed to utilize replicas to improve a system's performance, often requiring expensive coordination or sacrificing consistency. In this paper, we present SCAR, a new distributed and replicated in-memory database that allows serializable transactions to read from backup replicas with minimal coordination. SCAR works by assigning logical timestamps to database records so that a transaction can safely read from a backup replica without coordinating with the primary replica, because the records cannot be changed up to a certain logical time. In addition, we propose two optimization techniques, timestamp synchronization and parallel locking and validation, to further reduce coordination. We show that SCAR outperforms systems with conventional concurrency control algorithms and replication strategies by up to a factor of 2 on three popular benchmarks. We also demonstrate that SCAR achieves higher throughput by running under reduced isolation levels and detects concurrency anomalies in real time.