Clock-RSM: Low-Latency Inter-Datacenter State Machine Replication Using Loosely Synchronized Physical Clocks

被引:18
作者
Du, Jiaqing [1 ]
Sciascia, Daniele [2 ]
Elnikety, Sameh [3 ]
Zwaenepoel, Willy [1 ]
Pedone, Fernando [2 ]
机构
[1] Ecole Polytech Fed Lausanne, CH-1015 Lausanne, Switzerland
[2] Univ Lugano USI, Lugano, Switzerland
[3] Microsoft Res, Redmond, WA USA
来源
2014 44TH ANNUAL IEEE/IFIP INTERNATIONAL CONFERENCE ON DEPENDABLE SYSTEMS AND NETWORKS (DSN) | 2014年
关键词
D O I
10.1109/DSN.2014.42
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
This paper proposes Clock-RSM, a new state machine replication protocol that uses loosely synchronized physical clocks to totally order commands for geo-replicated services. Clock-RSM assumes realistic non-uniform latencies among replicas located at different data centers. It provides low-latency linearizable replication by overlapping 1) logging a command at a majority of replicas, 2) determining the stable order of the command from the farthest replica, and 3) notifying the commit of the command to all replicas. We evaluate Clock-RSM analytically and derive the expected command replication latency. We also evaluate the protocol experimentally using a geo-replicated key-value store deployed across multiple Amazon EC2 data centers.
引用
收藏
页码:343 / 354
页数:12
相关论文
共 20 条
[1]  
Baker J., 2011, CIDR
[2]  
BERENSON H, 1995, SIGMOD
[3]  
Bernstein P. A., 1986, Concurrency Control and Recovery in Database Systems
[4]  
Calder B. e. a., 2011, SOSP
[5]  
Corbett James C., 2012, OSDI
[6]  
Du J., 2014, TECHNICAL REPORT
[7]  
Herlihy Maurice P, 1990, TOPLAS
[8]  
Jain S., 2013, SIGCOMM
[9]  
Kemme B., 2000, VLDB
[10]  
Kraska T., 2013, EUROSYS