Strong Consistency at Scale

Today’s online services must meet strict availability and performance requirements. State machine replication, one of the most fundamental approaches to increasing the availability of services without sacrificing strong consistency, provides configurable availability but limited performance scalability. Scalable State Machine Replication (S-SMR) achieves scalable performance by partitioning the service state and coordinating the ordering and execution of commands. While S-SMR scales the performance of single-partition commands with the number of deployed partitions, replica coordination needed by multipartition commands introduces an overhead in the execution of multi-partition commands. In the paper, we review Scalable State Machine Replication and quantify the overhead due to replica coordination in different scenarios. In brief, we show that performance overhead is affected by the number of partitions involved in multi-partition commands and data locality.