SyncGC

Abstract

Data-center applications running on distributed databases often suffer from unexpectedly high response time fluctuation which is caused by long tail latency. In this paper, we find that long tail latency of user writes is mainly created by the interference with garbage collection (GC) tasks running in various system layers. In order to address the tail latency problem, we propose a synchronized garbage collection technique, called SyncGC. By scheduling multiple GC instances to execute in sync with each other in an overlapped manner, SyncGC prevents user requests from being interfered with GC instances, thereby minimizing their negative impacts on tail latency. Our experimental results with Cassandra show that SyncGC reduces the 99.99th-percentile tail latency and the maximum latency by 35% and 37%, on average, respectively.