一句话：提出 QoS 与不确定性感知调度器 Aquatope，为多阶段 Serverless 工作流预热容器并按函数分配资源，显著降低违规与成本。
多阶段 Serverless 工作流由大量计算与 I/O 阶段组成，虽具备细粒度扩展与模块化开发优势，却比简单 Serverless 函数更容易出现性能不佳、资源低效和高成本。Aquatope 是面向端到端 Serverless 工作流的 QoS 与不确定性感知资源调度器，显式考虑 FaaS 平台固有不确定性，利用一组可扩展且经验证的贝叶斯模型，在函数调用前创建预热容器，并以函数粒度分配合适资源，以在满足复杂工作流端到端 QoS 的同时最小化资源成本。在多种分析与交互式多阶段 Serverless 负载上，Aquatope 优于已有系统：相比其他满足 QoS 的方法，QoS 违规减少 5 倍，成本平均降低 34%，最高降低 52%。
Multi-stage serverless applications, i.e., workflows with many computation and I/O stages, are becoming increasingly representative of FaaS platforms. Despite their advantages in terms of fine-grained scalability and modular development, these applications are subject to suboptimal performance, resource inefficiency, and high costs to a larger degree than previous simple serverless functions. We present Aquatope, a QoS-and-uncertainty-aware resource scheduler for end-to-end serverless workflows that takes into account the inherent uncertainty present in FaaS platforms, and improves performance predictability and resource efficiency. Aquatope uses a set of scalable and validated Bayesian models to create pre-warmed containers ahead of function invocations, and to allocate appropriate resources at function granularity to meet a complex workflow's end-to-end QoS, while minimizing resource cost. Across a diverse set of analytics and interactive multi-stage serverless workloads, Aquatope significantly outperforms prior systems, reducing QoS violations by 5X, and cost by 34% on average and up to 52% compared to other QoS-meeting methods.
@inproceedings{ZhouZD23,
title = {AQUATOPE: QoS-and-Uncertainty-Aware Resource Management for Multi-stage Serverless Workflows},
author = {Zhuangzhuang Zhou and Yanqi Zhang and Christina Delimitrou},
booktitle = {ASPLOS},
year = {2022},
doi = {10.1145/3567955.3567960}
}