Cutting tail latency in cloud data stores via adaptive replica selection

Опубликовано 22 июня 2016, 1:18

Achieving predictable performance is critical for many distributed applications, yet difficult to achieve due to many factors that skew the tail of the latency distribution even in well-provisioned systems. In this talk, we will present the fundamental challenges involved in designing a replica selection scheme that is robust in the face of performance fluctuations across servers. We will then present the design and implementation of an adaptive replica selection mechanism, C3, that is robust to performance variability in the environment. We will also discuss C3’s effectiveness in reducing the latency tail and improving throughput through results of performance evaluations conducted on Amazon EC2 and through simulations. An implementation of C3 inside Cassandra improved tail latencies by factors exceeding 3x, while also improving throughput by up to 50%.