μ Suite: A Benchmark Suite for Microservices

TLDR

Modern OLDI applications have evolved into microservices that must meet sub‑millisecond RPC latency targets, making sub‑millisecond OS and network overheads dominant, yet existing benchmarks are monolithic and too slow to study these effects. The study aims to characterize how OS and network overheads affect microservice latency. The authors developed μSuite, a suite of four microservice‑based OLDI services—image similarity search, key‑value store routing, set algebra on posting lists, and recommender systems—to enable this study. The characterization shows that optimal OS/network settings depend on load in a complex way, and that suboptimal OS scheduler choices can increase microservice tail latency by up to about 87 %.

Abstract

Modern On-Line Data Intensive (OLDI) applications have evolved from monolithic systems to instead comprise numerous, distributed microservices interacting via Remote Procedure Calls (RPCs). Microservices face single-digit millisecond RPC latency goals (implying sub-ms medians)-much tighter than their monolithic ancestors that must meet ≥ 100 ms latency targets. Sub-ms-scale OS/network overheads that were once insignificant for such monoliths can now come to dominate in the sub-msscale microservice regime. It is therefore vital to characterize the influence of OSand network-based effects on microservices. Unfortunately, widely-used academic data center benchmark suites are unsuitable to aid this characterization as they (1) use monolithic rather than microservice architectures, and (2) largely have request service times ≥ 100 ms. In this paper, we investigate how OS and network overheads impact microservice median and tail latency by developing a complete suite of microservices called μSuite that we use to facilitate our study. μSuite comprises four OLDI services composed of microservices: image similarity search, protocol routing for key-value stores, set algebra on posting lists for document search, and recommender systems. Our characterization reveals that the relationship between optimal OS/network parameters and service load is complex. Our primary finding is that non-optimal OS scheduler decisions can degrade microservice tail latency by up to ~87%.