TY - GEN
T1 - Hill-Cache
T2 - 40th IEEE International Conference on Data Engineering, ICDE 2024
AU - Li, Yunfan
AU - Hu, Huiqi
AU - Lei, Chaojing
AU - Zhou, Xuan
AU - Qian, Weining
N1 - Publisher Copyright:
© 2024 IEEE.
PY - 2024
Y1 - 2024
N2 - Cache replacement policies are essential for maximizing application performance. Policies such as LRFU, which incorporate both recency and frequency, have shown efficacy in improving hit rates in many studies. In this paper, we theoretically investigated how parameters impact hit rates in LRFU and discovered two distinct features named unimodality and correlation. Drawing on our understanding, we formulated Hill-Cache. Hill-Cache provides a holistic approach to cache. It incorporates recency and frequency and employs a hill-climbing algorithm for adaptability. Additionally, it improves churn resistance through quick demotion, the approach that the most recent research suggests. By overcoming the limitations of LRFU, including high maintenance overhead and dependency on parameters, Hill-Cache distinguishes itself as a new cache method. Hill-Cache is well-suited for situations where there is an efficiency gap between performance devices and capacity devices, including data caching applications and database systems. Our evaluations, employing 36 real-world traces and various sophisticated policies across diverse cache sizes, demonstrated the superior performance of Hill-Cache. It reduces the average miss rate by 11.97% compared to LRU and outperforms other advanced cache policies such as ARC, LIRS, DLIRS, CACHEUS, and S3FIFO. We incorporated Hill-Cache into Memcached and RocksDB, significantly improving performance metrics such as throughput and latency.
AB - Cache replacement policies are essential for maximizing application performance. Policies such as LRFU, which incorporate both recency and frequency, have shown efficacy in improving hit rates in many studies. In this paper, we theoretically investigated how parameters impact hit rates in LRFU and discovered two distinct features named unimodality and correlation. Drawing on our understanding, we formulated Hill-Cache. Hill-Cache provides a holistic approach to cache. It incorporates recency and frequency and employs a hill-climbing algorithm for adaptability. Additionally, it improves churn resistance through quick demotion, the approach that the most recent research suggests. By overcoming the limitations of LRFU, including high maintenance overhead and dependency on parameters, Hill-Cache distinguishes itself as a new cache method. Hill-Cache is well-suited for situations where there is an efficiency gap between performance devices and capacity devices, including data caching applications and database systems. Our evaluations, employing 36 real-world traces and various sophisticated policies across diverse cache sizes, demonstrated the superior performance of Hill-Cache. It reduces the average miss rate by 11.97% compared to LRU and outperforms other advanced cache policies such as ARC, LIRS, DLIRS, CACHEUS, and S3FIFO. We incorporated Hill-Cache into Memcached and RocksDB, significantly improving performance metrics such as throughput and latency.
KW - cache replacement policy
KW - data caching system
KW - hill-climbing algorithm
KW - key-value storage
UR - https://www.scopus.com/pages/publications/85200520021
U2 - 10.1109/ICDE60146.2024.00302
DO - 10.1109/ICDE60146.2024.00302
M3 - 会议稿件
AN - SCOPUS:85200520021
T3 - Proceedings - International Conference on Data Engineering
SP - 3947
EP - 3960
BT - Proceedings - 2024 IEEE 40th International Conference on Data Engineering, ICDE 2024
PB - IEEE Computer Society
Y2 - 13 May 2024 through 17 May 2024
ER -