一句话：通过在 OS 内核中主动压缩冷内存页，以软件方式构建远内存层，显著降低仓库级计算机的内存总拥有成本。
内存需求持续增长与工艺微缩放缓，使仓库级计算机（WSC）的总拥有成本（TCO）面临严峻挑战。一种降低内存 TCO 的思路是引入更便宜但更慢的“远内存”层来存放不常访问的冷数据，但这会带来负载多样性与数据更替的动态响应、容量搁浅以及遗留（brownfield）部署等新问题。本文提出一种新颖的软件定义远内存方法：主动压缩冷内存页，从而在软件层面有效构造远内存层。端到端系统设计涵盖定义性能服务级目标（SLO）的新方法、在满足 SLO 前提下识别冷内存页的机制，以及 OS 内核与节点代理中的具体实现；同时设计基于学习的自动调参，无需人工介入即可周期性适应集群级变化。该系统自 2016 年起已在谷歌 WSC 上成功部署，服务数千个生产服务，内存成本降低 67% 以上，访问速度约 6 微秒，平均可容纳 20% 的不常访问数据，在仓库规模上带来显著的 TCO 节省。
Increasing memory demand and slowdown in technology scaling pose important challenges to total cost of ownership (TCO) of warehouse-scale computers (WSCs). One promising idea to reduce the memory TCO is to add a cheaper, but slower, "far memory" tier and use it to store infrequently accessed (or cold) data. However, introducing a far memory tier brings new challenges around dynamically responding to workload diversity and churn, minimizing stranding of capacity, and addressing brownfield (legacy) deployments. We present a novel software-defined approach to far memory that proactively compresses cold memory pages to effectively create a far memory tier in software. Our end-to-end system design encompasses new methods to define performance service-level objectives (SLOs), a mechanism to identify cold memory pages while meeting the SLO, and our implementation in the OS kernel and node agent. Additionally, we design learning-based autotuning to periodically adapt our design to fleet-wide changes without a human in the loop. Our system has been successfully deployed across Google's WSC since 2016, serving thousands of production services. Our software-defined far memory is significantly cheaper (67% or higher memory cost reduction) at relatively good access speeds (6us) and allows us to store a significant fraction of infrequently accessed data (on average, 20%), translating to significant TCO savings at warehouse scale.
@inproceedings{paperbot4918,
title = {Software-Defined Far Memory in Warehouse-Scale Computers},
author = {Andres Lagar-Cavilla and Junwhan Ahn and Suleiman Souhlal and Neha Agarwal and Radoslaw Burny and Shakeel Butt and Jichuan Chang and Ashwin Chaugule and Nan Deng and Junaid Shahid and Greg Thelen and Kamil Adam Yurtsever and Yu Zhao and Parthasarathy Ranganathan},
booktitle = {ASPLOS},
year = {2019},
doi = {10.1145/3297858.3304053}
}