一句话：提出 Thesios，融合多磁盘降采样 I/O 轨迹，合成高精度反事实全分辨率 I/O 轨迹。
大规模分布式存储系统设计需要代表性的 I/O 活动模型，尤其需要反事实“what-if”分析来评估新存储策略或硬件，但生产环境 A/B 实验成本高且有风险。Thesios 通过组合来自多台存储服务器、多块磁盘的降采样 I/O 轨迹，合成假设场景下的全分辨率 I/O 轨迹。作者利用 Google 已有常规采样轨迹进行评估，合成轨迹在读写请求数上达到 95–99.5% 准确率，利用率 90–97%，读延迟 80–99.8%。论文通过四个案例研究展示其能力：改变磁盘利用率、满度和容量，评估新数据放置策略，分析降低 RPM 对功耗与性能的影响，以及理解增大缓冲区缓存对存储服务器的影响。
Representative modeling of I/O activity is crucial when designing large-scale distributed storage systems. Particularly important use cases are counterfactual "what-if" analyses that assess the impact of anticipated or hypothetical new storage policies or hardware prior to deployment. We propose Thesios, a methodology to accurately synthesize such hypothetical full-resolution I/O traces by carefully combining down-sampled I/O traces collected from multiple disks attached to multiple storage servers. Applying this approach to real-world traces that are already routinely sampled at Google, we show that our synthesized traces achieve 95--99.5% accuracy in read/write request numbers, 90--97% accuracy in utilization, and 80--99.8% accuracy in read latency compared to metrics collected from actual disks. We demonstrate how Thesios enables diverse counterfactual I/O trace synthesis and analyses of hypothetical policy, hardware, and server changes through four case studies: (1) studying the effects of changing disk's utilization, fullness, and capacity, (2) evaluating new data placement policy, (3) analyzing the impact on power and performance of deploying disks with reduced rotations-per-minute (RPM), and (4) understanding the impact of increased buffer cache size on a storage server. Without Thesios, such counterfactual analyses would require costly and potentially risky A/B experiments in production.
@inproceedings{PhothilimthanaK24,
title = {Thesios: Synthesizing Accurate Counterfactual I/O Traces from I/O Samples},
author = {Phitchaya Mangpo Phothilimthana and Saurabh Kadekodi and Soroush Ghodrati and Selene Moon and Martin Maas},
booktitle = {ASPLOS},
year = {2024},
doi = {10.1145/3620666.3651337}
}