SwapKV: A Hotness Aware In-Memory Key-Value Store for Hybrid Memory Systems

In-memory Key-Value (KV) stores are widely deployed in modern data centers. These systems generally use DRAM as their storage medium, causing huge hardware costs. The emerging persistent memory (PMEM) is a potential substitute for DRAM, which has a lower price and larger capacity, but lower access s...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on knowledge and data engineering 2023-01, Vol.35 (1), p.917-930
Hauptverfasser: Cui, Lixiao, He, Kewen, Li, Yusen, Li, Peng, Zhang, Jiachen, Wang, Gang, Liu, Xiaoguang
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 930
container_issue 1
container_start_page 917
container_title IEEE transactions on knowledge and data engineering
container_volume 35
creator Cui, Lixiao
He, Kewen
Li, Yusen
Li, Peng
Zhang, Jiachen
Wang, Gang
Liu, Xiaoguang
description In-memory Key-Value (KV) stores are widely deployed in modern data centers. These systems generally use DRAM as their storage medium, causing huge hardware costs. The emerging persistent memory (PMEM) is a potential substitute for DRAM, which has a lower price and larger capacity, but lower access speed and bandwidth. Many prior studies strive to build hybrid memory systems to retain both the advantages of DRAM and PMEM. However, they are either application agnostic or simply take DRAM as a cache, which are both not efficient for in-memory KV stores. In this paper, we propose SwapKV, a well-designed in-memory KV store for hybrid DRAM-PMEM system. SwapKV has several promising properties. First, SwapKV combines DRAM and PMEM to a uniform memory pool and only stores one copy of data, which maximizes capacity utilization. Second, SwapKV maps all writing operations to DRAM and migrates data to PMEM with large blocks asynchronously, which mitigates the intrinsic inefficiency of PMEM for writing operations. Third, SwapKV maintains the hot data in DRAM through an efficient hotness filtering and data swapping mechanism, which ensures high system throughput and responsiveness. We implement SwapKV and evaluate it under various workload patterns. The results demonstrate that SwapKV improves the throughput by 11 \!\sim\! ∼ 41 percent compared to the state-of-the-art alternatives.
doi_str_mv 10.1109/TKDE.2021.3077264
format Article
fullrecord <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_ieee_primary_9423547</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9423547</ieee_id><sourcerecordid>2747609811</sourcerecordid><originalsourceid>FETCH-LOGICAL-c336t-13cfc71f86a64b2d55d5b4880c0b012b82f31cebe527e9ec54578377e135b69a3</originalsourceid><addsrcrecordid>eNo9kE1PAjEQhhujiYj-AOOliedipx_brjeiKGQxHkCuzbbMJhBgsV1C9t-7BOJp3mSedyZ5CHkEPgDg-cu8eB8NBBcwkNwYkakr0gOtLROQw3WXuQKmpDK35C6lNefcGgs9UsyO5b5YvNIhHdfNDlOiw2MZkU527Au3dWxpgS1blJsD0llTd5uqjnTc-rha0gsxa1OD23RPbqpyk_DhMvvk52M0fxuz6ffn5G04ZUHKrGEgQxUMVDYrM-XFUuul9spaHrjnILwVlYSAHrUwmGPQShsrjUGQ2md5Kfvk-Xx3H-vfA6bGretD3HUvnTDKZDy3AB0FZyrEOqWIldvH1baMrQPuTs7cyZk7OXMXZ13n6dxZIeI_nyshtTLyD_2tZeY</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2747609811</pqid></control><display><type>article</type><title>SwapKV: A Hotness Aware In-Memory Key-Value Store for Hybrid Memory Systems</title><source>IEEE Xplore</source><creator>Cui, Lixiao ; He, Kewen ; Li, Yusen ; Li, Peng ; Zhang, Jiachen ; Wang, Gang ; Liu, Xiaoguang</creator><creatorcontrib>Cui, Lixiao ; He, Kewen ; Li, Yusen ; Li, Peng ; Zhang, Jiachen ; Wang, Gang ; Liu, Xiaoguang</creatorcontrib><description><![CDATA[In-memory Key-Value (KV) stores are widely deployed in modern data centers. These systems generally use DRAM as their storage medium, causing huge hardware costs. The emerging persistent memory (PMEM) is a potential substitute for DRAM, which has a lower price and larger capacity, but lower access speed and bandwidth. Many prior studies strive to build hybrid memory systems to retain both the advantages of DRAM and PMEM. However, they are either application agnostic or simply take DRAM as a cache, which are both not efficient for in-memory KV stores. In this paper, we propose SwapKV, a well-designed in-memory KV store for hybrid DRAM-PMEM system. SwapKV has several promising properties. First, SwapKV combines DRAM and PMEM to a uniform memory pool and only stores one copy of data, which maximizes capacity utilization. Second, SwapKV maps all writing operations to DRAM and migrates data to PMEM with large blocks asynchronously, which mitigates the intrinsic inefficiency of PMEM for writing operations. Third, SwapKV maintains the hot data in DRAM through an efficient hotness filtering and data swapping mechanism, which ensures high system throughput and responsiveness. We implement SwapKV and evaluate it under various workload patterns. The results demonstrate that SwapKV improves the throughput by 11 <inline-formula><tex-math notation="LaTeX">\!\sim\!</tex-math> <mml:math><mml:mrow><mml:mspace width="-0.166667em"/><mml:mo>∼</mml:mo><mml:mspace width="-0.166667em"/></mml:mrow></mml:math><inline-graphic xlink:href="cui-ieq1-3077264.gif"/> </inline-formula> 41 percent compared to the state-of-the-art alternatives.]]></description><identifier>ISSN: 1041-4347</identifier><identifier>EISSN: 1558-2191</identifier><identifier>DOI: 10.1109/TKDE.2021.3077264</identifier><identifier>CODEN: ITKEEH</identifier><language>eng</language><publisher>New York: IEEE</publisher><subject>Bandwidth ; Data centers ; Data storage ; Dynamic random access memory ; Hardware ; hybrid memory system ; Hybrid systems ; Indexes ; Key-value system ; Nonvolatile memory ; persistent memory ; Random access memory ; Throughput ; Writing</subject><ispartof>IEEE transactions on knowledge and data engineering, 2023-01, Vol.35 (1), p.917-930</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c336t-13cfc71f86a64b2d55d5b4880c0b012b82f31cebe527e9ec54578377e135b69a3</citedby><cites>FETCH-LOGICAL-c336t-13cfc71f86a64b2d55d5b4880c0b012b82f31cebe527e9ec54578377e135b69a3</cites><orcidid>0000-0002-4726-4615 ; 0000-0002-4017-0974 ; 0000-0003-0387-2501 ; 0000-0002-9010-3278 ; 0000-0001-6623-350X</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9423547$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,780,784,796,27915,27916,54749</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/9423547$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Cui, Lixiao</creatorcontrib><creatorcontrib>He, Kewen</creatorcontrib><creatorcontrib>Li, Yusen</creatorcontrib><creatorcontrib>Li, Peng</creatorcontrib><creatorcontrib>Zhang, Jiachen</creatorcontrib><creatorcontrib>Wang, Gang</creatorcontrib><creatorcontrib>Liu, Xiaoguang</creatorcontrib><title>SwapKV: A Hotness Aware In-Memory Key-Value Store for Hybrid Memory Systems</title><title>IEEE transactions on knowledge and data engineering</title><addtitle>TKDE</addtitle><description><![CDATA[In-memory Key-Value (KV) stores are widely deployed in modern data centers. These systems generally use DRAM as their storage medium, causing huge hardware costs. The emerging persistent memory (PMEM) is a potential substitute for DRAM, which has a lower price and larger capacity, but lower access speed and bandwidth. Many prior studies strive to build hybrid memory systems to retain both the advantages of DRAM and PMEM. However, they are either application agnostic or simply take DRAM as a cache, which are both not efficient for in-memory KV stores. In this paper, we propose SwapKV, a well-designed in-memory KV store for hybrid DRAM-PMEM system. SwapKV has several promising properties. First, SwapKV combines DRAM and PMEM to a uniform memory pool and only stores one copy of data, which maximizes capacity utilization. Second, SwapKV maps all writing operations to DRAM and migrates data to PMEM with large blocks asynchronously, which mitigates the intrinsic inefficiency of PMEM for writing operations. Third, SwapKV maintains the hot data in DRAM through an efficient hotness filtering and data swapping mechanism, which ensures high system throughput and responsiveness. We implement SwapKV and evaluate it under various workload patterns. The results demonstrate that SwapKV improves the throughput by 11 <inline-formula><tex-math notation="LaTeX">\!\sim\!</tex-math> <mml:math><mml:mrow><mml:mspace width="-0.166667em"/><mml:mo>∼</mml:mo><mml:mspace width="-0.166667em"/></mml:mrow></mml:math><inline-graphic xlink:href="cui-ieq1-3077264.gif"/> </inline-formula> 41 percent compared to the state-of-the-art alternatives.]]></description><subject>Bandwidth</subject><subject>Data centers</subject><subject>Data storage</subject><subject>Dynamic random access memory</subject><subject>Hardware</subject><subject>hybrid memory system</subject><subject>Hybrid systems</subject><subject>Indexes</subject><subject>Key-value system</subject><subject>Nonvolatile memory</subject><subject>persistent memory</subject><subject>Random access memory</subject><subject>Throughput</subject><subject>Writing</subject><issn>1041-4347</issn><issn>1558-2191</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNo9kE1PAjEQhhujiYj-AOOliedipx_brjeiKGQxHkCuzbbMJhBgsV1C9t-7BOJp3mSedyZ5CHkEPgDg-cu8eB8NBBcwkNwYkakr0gOtLROQw3WXuQKmpDK35C6lNefcGgs9UsyO5b5YvNIhHdfNDlOiw2MZkU527Au3dWxpgS1blJsD0llTd5uqjnTc-rha0gsxa1OD23RPbqpyk_DhMvvk52M0fxuz6ffn5G04ZUHKrGEgQxUMVDYrM-XFUuul9spaHrjnILwVlYSAHrUwmGPQShsrjUGQ2md5Kfvk-Xx3H-vfA6bGretD3HUvnTDKZDy3AB0FZyrEOqWIldvH1baMrQPuTs7cyZk7OXMXZ13n6dxZIeI_nyshtTLyD_2tZeY</recordid><startdate>20230101</startdate><enddate>20230101</enddate><creator>Cui, Lixiao</creator><creator>He, Kewen</creator><creator>Li, Yusen</creator><creator>Li, Peng</creator><creator>Zhang, Jiachen</creator><creator>Wang, Gang</creator><creator>Liu, Xiaoguang</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><orcidid>https://orcid.org/0000-0002-4726-4615</orcidid><orcidid>https://orcid.org/0000-0002-4017-0974</orcidid><orcidid>https://orcid.org/0000-0003-0387-2501</orcidid><orcidid>https://orcid.org/0000-0002-9010-3278</orcidid><orcidid>https://orcid.org/0000-0001-6623-350X</orcidid></search><sort><creationdate>20230101</creationdate><title>SwapKV: A Hotness Aware In-Memory Key-Value Store for Hybrid Memory Systems</title><author>Cui, Lixiao ; He, Kewen ; Li, Yusen ; Li, Peng ; Zhang, Jiachen ; Wang, Gang ; Liu, Xiaoguang</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c336t-13cfc71f86a64b2d55d5b4880c0b012b82f31cebe527e9ec54578377e135b69a3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Bandwidth</topic><topic>Data centers</topic><topic>Data storage</topic><topic>Dynamic random access memory</topic><topic>Hardware</topic><topic>hybrid memory system</topic><topic>Hybrid systems</topic><topic>Indexes</topic><topic>Key-value system</topic><topic>Nonvolatile memory</topic><topic>persistent memory</topic><topic>Random access memory</topic><topic>Throughput</topic><topic>Writing</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Cui, Lixiao</creatorcontrib><creatorcontrib>He, Kewen</creatorcontrib><creatorcontrib>Li, Yusen</creatorcontrib><creatorcontrib>Li, Peng</creatorcontrib><creatorcontrib>Zhang, Jiachen</creatorcontrib><creatorcontrib>Wang, Gang</creatorcontrib><creatorcontrib>Liu, Xiaoguang</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005–Present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998–Present</collection><collection>IEEE Xplore</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>IEEE transactions on knowledge and data engineering</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Cui, Lixiao</au><au>He, Kewen</au><au>Li, Yusen</au><au>Li, Peng</au><au>Zhang, Jiachen</au><au>Wang, Gang</au><au>Liu, Xiaoguang</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>SwapKV: A Hotness Aware In-Memory Key-Value Store for Hybrid Memory Systems</atitle><jtitle>IEEE transactions on knowledge and data engineering</jtitle><stitle>TKDE</stitle><date>2023-01-01</date><risdate>2023</risdate><volume>35</volume><issue>1</issue><spage>917</spage><epage>930</epage><pages>917-930</pages><issn>1041-4347</issn><eissn>1558-2191</eissn><coden>ITKEEH</coden><abstract><![CDATA[In-memory Key-Value (KV) stores are widely deployed in modern data centers. These systems generally use DRAM as their storage medium, causing huge hardware costs. The emerging persistent memory (PMEM) is a potential substitute for DRAM, which has a lower price and larger capacity, but lower access speed and bandwidth. Many prior studies strive to build hybrid memory systems to retain both the advantages of DRAM and PMEM. However, they are either application agnostic or simply take DRAM as a cache, which are both not efficient for in-memory KV stores. In this paper, we propose SwapKV, a well-designed in-memory KV store for hybrid DRAM-PMEM system. SwapKV has several promising properties. First, SwapKV combines DRAM and PMEM to a uniform memory pool and only stores one copy of data, which maximizes capacity utilization. Second, SwapKV maps all writing operations to DRAM and migrates data to PMEM with large blocks asynchronously, which mitigates the intrinsic inefficiency of PMEM for writing operations. Third, SwapKV maintains the hot data in DRAM through an efficient hotness filtering and data swapping mechanism, which ensures high system throughput and responsiveness. We implement SwapKV and evaluate it under various workload patterns. The results demonstrate that SwapKV improves the throughput by 11 <inline-formula><tex-math notation="LaTeX">\!\sim\!</tex-math> <mml:math><mml:mrow><mml:mspace width="-0.166667em"/><mml:mo>∼</mml:mo><mml:mspace width="-0.166667em"/></mml:mrow></mml:math><inline-graphic xlink:href="cui-ieq1-3077264.gif"/> </inline-formula> 41 percent compared to the state-of-the-art alternatives.]]></abstract><cop>New York</cop><pub>IEEE</pub><doi>10.1109/TKDE.2021.3077264</doi><tpages>14</tpages><orcidid>https://orcid.org/0000-0002-4726-4615</orcidid><orcidid>https://orcid.org/0000-0002-4017-0974</orcidid><orcidid>https://orcid.org/0000-0003-0387-2501</orcidid><orcidid>https://orcid.org/0000-0002-9010-3278</orcidid><orcidid>https://orcid.org/0000-0001-6623-350X</orcidid><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1041-4347
ispartof IEEE transactions on knowledge and data engineering, 2023-01, Vol.35 (1), p.917-930
issn 1041-4347
1558-2191
language eng
recordid cdi_ieee_primary_9423547
source IEEE Xplore
subjects Bandwidth
Data centers
Data storage
Dynamic random access memory
Hardware
hybrid memory system
Hybrid systems
Indexes
Key-value system
Nonvolatile memory
persistent memory
Random access memory
Throughput
Writing
title SwapKV: A Hotness Aware In-Memory Key-Value Store for Hybrid Memory Systems
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-14T19%3A51%3A03IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=SwapKV:%20A%20Hotness%20Aware%20In-Memory%20Key-Value%20Store%20for%20Hybrid%20Memory%20Systems&rft.jtitle=IEEE%20transactions%20on%20knowledge%20and%20data%20engineering&rft.au=Cui,%20Lixiao&rft.date=2023-01-01&rft.volume=35&rft.issue=1&rft.spage=917&rft.epage=930&rft.pages=917-930&rft.issn=1041-4347&rft.eissn=1558-2191&rft.coden=ITKEEH&rft_id=info:doi/10.1109/TKDE.2021.3077264&rft_dat=%3Cproquest_RIE%3E2747609811%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2747609811&rft_id=info:pmid/&rft_ieee_id=9423547&rfr_iscdi=true