Maximizing I/O Throughput and Minimizing Performance Variation via Reinforcement Learning Based I/O Merging for SSDs
Research output: Journal Publications and Reviews (RGC: 21, 22, 62) › 21_Publication in refereed journal › peer-review
Author(s)
Related Research Unit(s)
Detail(s)
Original language | English |
---|---|
Article number | 8823009 |
Pages (from-to) | 72-86 |
Journal / Publication | IEEE Transactions on Computers |
Volume | 69 |
Issue number | 1 |
Online published | 3 Sep 2019 |
Publication status | Published - Jan 2020 |
Link(s)
Abstract
Merging technique is widely adopted by I/O schedulers to maximize system I/O throughput. However, I/O merging could increase the latency of individual I/O, thus incurring prolonged I/O latencies and enlarged performance variations. Even with better system throughput, higher worst-case latency experienced by some requests could block the SSD storage system, which violates the QoS (Quality of Service) requirement. In order to improve QoS performance while providing higher I/O throughput, this paper proposes a reinforcement learning based I/O merging approach. Through learning the characteristic of various I/O patterns, the proposed approach makes merging decisions adaptively based on different I/O workloads. Evaluation results show that the proposed scheme is capable of reducing the standard deviation of I/O latency by 19.1 percent on average, worst-case latency by 7.3-60.9 percent at the 99.9th percentile compared with the latest I/O merging scheme, while maximizing system throughput.
Research Area(s)
- I/O scheduler, Merging technique, performance variation, reinforcement learning, throughput, worst-case latency
Citation Format(s)
Maximizing I/O Throughput and Minimizing Performance Variation via Reinforcement Learning Based I/O Merging for SSDs. / Wu, Chao; Ji, Cheng; Li, Qiao; Gao, Congming; Pan, Riwei; Fu, Chenchen; Shi, Liang; Xue, Chun Jason.
In: IEEE Transactions on Computers, Vol. 69, No. 1, 8823009, 01.2020, p. 72-86.Research output: Journal Publications and Reviews (RGC: 21, 22, 62) › 21_Publication in refereed journal › peer-review