一句话：提出基于机器学习且感知 QoS 的云微服务资源管理系统 Sinan，缓解服务依赖引发的级联 QoS 违规。
云应用正从大型单体服务转向大量松散耦合、职责专一的微服务。微服务在开发、部署、模块化与隔离方面具有优势，但服务之间的依赖会引入背压效应，并导致级联的 QoS 违规，从而使资源管理变得复杂。针对这一问题，Sinan 提出一种基于机器学习（ML）且感知 QoS 的资源管理方案：以机器学习方法支撑资源管理决策，并把端到端的 QoS 目标纳入考量，以缓解背压传播与级联违规。所提供的摘要仅描述了动机与问题背景，未披露具体的技术机制、实现细节与实验结果。
Cloud applications are increasingly shifting from large monolithic services, to large numbers of loosely-coupled, specialized microservices. Despite their advantages in terms of facilitating development, deployment, modularity, and isolation, microservices complicate resource management, as dependencies between them introduce backpressure effects and cascading QoS violations.
DOI 原文 ·
@inproceedings{paperbot6092,
title = {Sinan: ML-based and QoS-aware resource management for cloud microservices},
author = {Yanqi Zhang and Weizhe Hua and Zhuangzhuang Zhou and G. Edward Suh and Christina Delimitrou},
booktitle = {ASPLOS},
year = {2021},
doi = {10.1145/3445814.3446693}
}