TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic

Disaggregated storage (DS) based on remote direct memory access (RDMA) network decouples compute and storage resources, thereby significantly improving resource utilization. While building key-value (KV) stores on DS benefits from these merits, existing fast KV stores suffer from network bandwidth c...

Full description

Saved in:
Bibliographic Details
Main Authors: Ling Zhan, Kai Lu, Yiqin Xiong, Jiguang Wan, Zixuan Yang
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10752495/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850130187134435328
author Ling Zhan
Kai Lu
Yiqin Xiong
Jiguang Wan
Zixuan Yang
author_facet Ling Zhan
Kai Lu
Yiqin Xiong
Jiguang Wan
Zixuan Yang
author_sort Ling Zhan
collection DOAJ
description Disaggregated storage (DS) based on remote direct memory access (RDMA) network decouples compute and storage resources, thereby significantly improving resource utilization. While building key-value (KV) stores on DS benefits from these merits, existing fast KV stores suffer from network bandwidth contention and high latency under DS due to the non-negligible network amplification and high-overhead I/O stack. In this paper, we propose TrickleKV, a high-performance persistent KV store designed for DS. TrickleKV reduces network amplification and latency in three approaches: 1) TrickleKV proposes an efficient storage-side data filtering mechanism and a two-level cache structure with different granularities to reduce network traffic in the read process. 2) TrickleKV presents an efficient write buffer structure that includes asynchronous flushing and queue scheduling mechanisms to reduce network traffic in the write process. 3) TrickleKV designs a read-write decoupled user-space I/O stack and lightweight storage space management to reduce access latency. Evaluation results show that TrickleKV achieves <inline-formula> <tex-math notation="LaTeX">$1.2\times $ </tex-math></inline-formula>&#x2013;<inline-formula> <tex-math notation="LaTeX">$7\times $ </tex-math></inline-formula> higher throughput and 30%-<inline-formula> <tex-math notation="LaTeX">$7.4\times $ </tex-math></inline-formula> lower latency compared to state-of-the-art KV stores under DS.
format Article
id doaj-art-fd9118c1f88b49dab8bfbdcbcf1263ee
institution OA Journals
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-fd9118c1f88b49dab8bfbdcbcf1263ee2025-08-20T02:32:45ZengIEEEIEEE Access2169-35362024-01-011216759616761210.1109/ACCESS.2024.349688010752495TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network TrafficLing Zhan0Kai Lu1https://orcid.org/0000-0002-7757-4083Yiqin Xiong2https://orcid.org/0009-0003-2113-7014Jiguang Wan3https://orcid.org/0000-0002-4160-9475Zixuan Yang4Faculty of Information Science and Technology, Wenhua College, Wuhan, ChinaWuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan, ChinaWuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan, ChinaWuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan, ChinaSchool of Big Data and Artificial Intelligence, Fujian Normal University, Fuzhou, ChinaDisaggregated storage (DS) based on remote direct memory access (RDMA) network decouples compute and storage resources, thereby significantly improving resource utilization. While building key-value (KV) stores on DS benefits from these merits, existing fast KV stores suffer from network bandwidth contention and high latency under DS due to the non-negligible network amplification and high-overhead I/O stack. In this paper, we propose TrickleKV, a high-performance persistent KV store designed for DS. TrickleKV reduces network amplification and latency in three approaches: 1) TrickleKV proposes an efficient storage-side data filtering mechanism and a two-level cache structure with different granularities to reduce network traffic in the read process. 2) TrickleKV presents an efficient write buffer structure that includes asynchronous flushing and queue scheduling mechanisms to reduce network traffic in the write process. 3) TrickleKV designs a read-write decoupled user-space I/O stack and lightweight storage space management to reduce access latency. Evaluation results show that TrickleKV achieves <inline-formula> <tex-math notation="LaTeX">$1.2\times $ </tex-math></inline-formula>&#x2013;<inline-formula> <tex-math notation="LaTeX">$7\times $ </tex-math></inline-formula> higher throughput and 30%-<inline-formula> <tex-math notation="LaTeX">$7.4\times $ </tex-math></inline-formula> lower latency compared to state-of-the-art KV stores under DS.https://ieeexplore.ieee.org/document/10752495/Disaggregated storagekey-value storeNVMe over fabricsremote direct memory access
spellingShingle Ling Zhan
Kai Lu
Yiqin Xiong
Jiguang Wan
Zixuan Yang
TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic
IEEE Access
Disaggregated storage
key-value store
NVMe over fabrics
remote direct memory access
title TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic
title_full TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic
title_fullStr TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic
title_full_unstemmed TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic
title_short TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage With Low Network Traffic
title_sort tricklekv a high performance key value store on disaggregated storage with low network traffic
topic Disaggregated storage
key-value store
NVMe over fabrics
remote direct memory access
url https://ieeexplore.ieee.org/document/10752495/
work_keys_str_mv AT lingzhan tricklekvahighperformancekeyvaluestoreondisaggregatedstoragewithlownetworktraffic
AT kailu tricklekvahighperformancekeyvaluestoreondisaggregatedstoragewithlownetworktraffic
AT yiqinxiong tricklekvahighperformancekeyvaluestoreondisaggregatedstoragewithlownetworktraffic
AT jiguangwan tricklekvahighperformancekeyvaluestoreondisaggregatedstoragewithlownetworktraffic
AT zixuanyang tricklekvahighperformancekeyvaluestoreondisaggregatedstoragewithlownetworktraffic