TY - JOUR
T1 - Maximizing I/O Throughput and Minimizing Performance Variation via Reinforcement Learning Based I/O Merging for SSDs
AU - Wu, Chao
AU - Ji, Cheng
AU - Li, Qiao
AU - Gao, Congming
AU - Pan, Riwei
AU - Fu, Chenchen
AU - Shi, Liang
AU - Xue, Chun Jason
PY - 2020/1
Y1 - 2020/1
N2 - 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.
AB - 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.
KW - I/O scheduler
KW - Merging technique
KW - performance variation
KW - reinforcement learning
KW - throughput
KW - worst-case latency
UR - http://www.scopus.com/inward/record.url?scp=85077797453&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85077797453&origin=recordpage
U2 - 10.1109/TC.2019.2938956
DO - 10.1109/TC.2019.2938956
M3 - RGC 21 - Publication in refereed journal
SN - 0018-9340
VL - 69
SP - 72
EP - 86
JO - IEEE Transactions on Computers
JF - IEEE Transactions on Computers
IS - 1
M1 - 8823009
ER -