Near-Optimal Stochastic Bin-Packing in Large Service Systems with Time-Varying Item Sizes

In modern computing systems, jobs' resource requirements often vary over time. Accounting for this temporal variability during job scheduling is essential for meeting performance goals. However, theoretical understanding on how to schedule jobs with time-varying resource requirements is limited...

Full description

Saved in:
Bibliographic Details
Published in:Performance evaluation review 2024-06, Vol.52 (1), p.93-94
Main Authors: Hong, Yige, Xie, Qiaomin, Wang, Weina
Format: Article
Language:English
Subjects:
Markov processes
Mathematics of computing
Network performance analysis
Network performance evaluation
Networks
Probability and statistics
Stochastic processes
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites cdi_FETCH-LOGICAL-a590-a52ebe18d22351322436f05d7bbce5a1f0761963508829ad7a22b57b76b622b33
container_end_page 94
container_issue 1
container_start_page 93
container_title Performance evaluation review
container_volume 52
creator Hong, Yige
Xie, Qiaomin
Wang, Weina
description In modern computing systems, jobs' resource requirements often vary over time. Accounting for this temporal variability during job scheduling is essential for meeting performance goals. However, theoretical understanding on how to schedule jobs with time-varying resource requirements is limited. Motivated by this gap, we propose a new setting of the stochastic bin-packing problem in service systems that allows for time-varying job resource requirements, also referred to as 'item sizes' in traditional bin-packing terms. In this setting, a job or 'item' must be dispatched to a server or 'bin' upon arrival. Its resource requirement varies over time as a Markov chain while in service. Our goal is to minimize the expected number of active servers, or 'non-empty bins', in the steady state. Under our problem formulation, we develop a job dispatch policy, named Join-Requesting-Server (JRS). We show that in the asymptotic regime where the job arrival rate scales large linearly with respect to a scaling factor r, JRS achieves an additive optimality gap of O(√r) in the objective value, where the optimal objective value is Θ(r). Our approach highlights a novel policy conversion framework that utilizes the solution of a single-server problem.
doi_str_mv 10.1145/3673660.3655070
format article
fullrecord <record><control><sourceid>acm_cross</sourceid><recordid>TN_cdi_crossref_primary_10_1145_3673660_3655070</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>3655070</sourcerecordid><originalsourceid>FETCH-LOGICAL-a590-a52ebe18d22351322436f05d7bbce5a1f0761963508829ad7a22b57b76b622b33</originalsourceid><addsrcrecordid>eNo9kF9LwzAUxYMoWKfPgk_5Atnyp0maRx06B8UJLYJPJUnTLbpuIylK_fR2rPpyz-Xecy7cHwC3BE8JSfmMCcmEwFMmOMcSn4GEcC6RSrP0HCSYCIa4UuoSXMX4gTGRlGQJeH9xOqDVofOt3sKi29uNjp238MHv0Ku2n363hn4Hcx3WDhYufHk7aB8710b47bsNLH3r0JsO_dG6HOaw8D8uXoOLRm-juxl1Asqnx3L-jPLVYjm_z5HmCg-FOuNIVlPKOGGUpkw0mNfSGOu4Jg2WgijBOM4yqnQtNaWGSyOFEUPH2ATMTmdt2McYXFMdwvBL6CuCqyOYagRTjWCGxN0poW37b_5b_gJ_g1z4</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Near-Optimal Stochastic Bin-Packing in Large Service Systems with Time-Varying Item Sizes</title><source>Association for Computing Machinery:Jisc Collections:ACM OPEN Journals 2023-2025 (reading list)</source><creator>Hong, Yige ; Xie, Qiaomin ; Wang, Weina</creator><creatorcontrib>Hong, Yige ; Xie, Qiaomin ; Wang, Weina</creatorcontrib><description>In modern computing systems, jobs' resource requirements often vary over time. Accounting for this temporal variability during job scheduling is essential for meeting performance goals. However, theoretical understanding on how to schedule jobs with time-varying resource requirements is limited. Motivated by this gap, we propose a new setting of the stochastic bin-packing problem in service systems that allows for time-varying job resource requirements, also referred to as 'item sizes' in traditional bin-packing terms. In this setting, a job or 'item' must be dispatched to a server or 'bin' upon arrival. Its resource requirement varies over time as a Markov chain while in service. Our goal is to minimize the expected number of active servers, or 'non-empty bins', in the steady state. Under our problem formulation, we develop a job dispatch policy, named Join-Requesting-Server (JRS). We show that in the asymptotic regime where the job arrival rate scales large linearly with respect to a scaling factor r, JRS achieves an additive optimality gap of O(√r) in the objective value, where the optimal objective value is Θ(r). Our approach highlights a novel policy conversion framework that utilizes the solution of a single-server problem.</description><identifier>ISSN: 0163-5999</identifier><identifier>EISSN: 1557-9484</identifier><identifier>DOI: 10.1145/3673660.3655070</identifier><language>eng</language><publisher>New York, NY, USA: ACM</publisher><subject>Markov processes ; Mathematics of computing ; Network performance analysis ; Network performance evaluation ; Networks ; Probability and statistics ; Stochastic processes</subject><ispartof>Performance evaluation review, 2024-06, Vol.52 (1), p.93-94</ispartof><rights>Owner/Author</rights><woscitedreferencessubscribed>false</woscitedreferencessubscribed><cites>FETCH-LOGICAL-a590-a52ebe18d22351322436f05d7bbce5a1f0761963508829ad7a22b57b76b622b33</cites><orcidid>0000-0001-8534-1063 ; 0000-0003-2834-6866 ; 0000-0001-6808-0156</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>315,786,790,27957,27958</link.rule.ids></links><search><creatorcontrib>Hong, Yige</creatorcontrib><creatorcontrib>Xie, Qiaomin</creatorcontrib><creatorcontrib>Wang, Weina</creatorcontrib><title>Near-Optimal Stochastic Bin-Packing in Large Service Systems with Time-Varying Item Sizes</title><title>Performance evaluation review</title><addtitle>ACM SIGMETRICS</addtitle><description>In modern computing systems, jobs' resource requirements often vary over time. Accounting for this temporal variability during job scheduling is essential for meeting performance goals. However, theoretical understanding on how to schedule jobs with time-varying resource requirements is limited. Motivated by this gap, we propose a new setting of the stochastic bin-packing problem in service systems that allows for time-varying job resource requirements, also referred to as 'item sizes' in traditional bin-packing terms. In this setting, a job or 'item' must be dispatched to a server or 'bin' upon arrival. Its resource requirement varies over time as a Markov chain while in service. Our goal is to minimize the expected number of active servers, or 'non-empty bins', in the steady state. Under our problem formulation, we develop a job dispatch policy, named Join-Requesting-Server (JRS). We show that in the asymptotic regime where the job arrival rate scales large linearly with respect to a scaling factor r, JRS achieves an additive optimality gap of O(√r) in the objective value, where the optimal objective value is Θ(r). Our approach highlights a novel policy conversion framework that utilizes the solution of a single-server problem.</description><subject>Markov processes</subject><subject>Mathematics of computing</subject><subject>Network performance analysis</subject><subject>Network performance evaluation</subject><subject>Networks</subject><subject>Probability and statistics</subject><subject>Stochastic processes</subject><issn>0163-5999</issn><issn>1557-9484</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><recordid>eNo9kF9LwzAUxYMoWKfPgk_5Atnyp0maRx06B8UJLYJPJUnTLbpuIylK_fR2rPpyz-Xecy7cHwC3BE8JSfmMCcmEwFMmOMcSn4GEcC6RSrP0HCSYCIa4UuoSXMX4gTGRlGQJeH9xOqDVofOt3sKi29uNjp238MHv0Ku2n363hn4Hcx3WDhYufHk7aB8710b47bsNLH3r0JsO_dG6HOaw8D8uXoOLRm-juxl1Asqnx3L-jPLVYjm_z5HmCg-FOuNIVlPKOGGUpkw0mNfSGOu4Jg2WgijBOM4yqnQtNaWGSyOFEUPH2ATMTmdt2McYXFMdwvBL6CuCqyOYagRTjWCGxN0poW37b_5b_gJ_g1z4</recordid><startdate>20240611</startdate><enddate>20240611</enddate><creator>Hong, Yige</creator><creator>Xie, Qiaomin</creator><creator>Wang, Weina</creator><general>ACM</general><scope>AAYXX</scope><scope>CITATION</scope><orcidid>https://orcid.org/0000-0001-8534-1063</orcidid><orcidid>https://orcid.org/0000-0003-2834-6866</orcidid><orcidid>https://orcid.org/0000-0001-6808-0156</orcidid></search><sort><creationdate>20240611</creationdate><title>Near-Optimal Stochastic Bin-Packing in Large Service Systems with Time-Varying Item Sizes</title><author>Hong, Yige ; Xie, Qiaomin ; Wang, Weina</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a590-a52ebe18d22351322436f05d7bbce5a1f0761963508829ad7a22b57b76b622b33</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Markov processes</topic><topic>Mathematics of computing</topic><topic>Network performance analysis</topic><topic>Network performance evaluation</topic><topic>Networks</topic><topic>Probability and statistics</topic><topic>Stochastic processes</topic><toplevel>online_resources</toplevel><creatorcontrib>Hong, Yige</creatorcontrib><creatorcontrib>Xie, Qiaomin</creatorcontrib><creatorcontrib>Wang, Weina</creatorcontrib><collection>CrossRef</collection><jtitle>Performance evaluation review</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Hong, Yige</au><au>Xie, Qiaomin</au><au>Wang, Weina</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Near-Optimal Stochastic Bin-Packing in Large Service Systems with Time-Varying Item Sizes</atitle><jtitle>Performance evaluation review</jtitle><stitle>ACM SIGMETRICS</stitle><date>2024-06-11</date><risdate>2024</risdate><volume>52</volume><issue>1</issue><spage>93</spage><epage>94</epage><pages>93-94</pages><issn>0163-5999</issn><eissn>1557-9484</eissn><abstract>In modern computing systems, jobs' resource requirements often vary over time. Accounting for this temporal variability during job scheduling is essential for meeting performance goals. However, theoretical understanding on how to schedule jobs with time-varying resource requirements is limited. Motivated by this gap, we propose a new setting of the stochastic bin-packing problem in service systems that allows for time-varying job resource requirements, also referred to as 'item sizes' in traditional bin-packing terms. In this setting, a job or 'item' must be dispatched to a server or 'bin' upon arrival. Its resource requirement varies over time as a Markov chain while in service. Our goal is to minimize the expected number of active servers, or 'non-empty bins', in the steady state. Under our problem formulation, we develop a job dispatch policy, named Join-Requesting-Server (JRS). We show that in the asymptotic regime where the job arrival rate scales large linearly with respect to a scaling factor r, JRS achieves an additive optimality gap of O(√r) in the objective value, where the optimal objective value is Θ(r). Our approach highlights a novel policy conversion framework that utilizes the solution of a single-server problem.</abstract><cop>New York, NY, USA</cop><pub>ACM</pub><doi>10.1145/3673660.3655070</doi><tpages>2</tpages><orcidid>https://orcid.org/0000-0001-8534-1063</orcidid><orcidid>https://orcid.org/0000-0003-2834-6866</orcidid><orcidid>https://orcid.org/0000-0001-6808-0156</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 0163-5999
ispartof Performance evaluation review, 2024-06, Vol.52 (1), p.93-94
issn 0163-5999
1557-9484
language eng
recordid cdi_crossref_primary_10_1145_3673660_3655070
source Association for Computing Machinery:Jisc Collections:ACM OPEN Journals 2023-2025 (reading list)
subjects Markov processes
Mathematics of computing
Network performance analysis
Network performance evaluation
Networks
Probability and statistics
Stochastic processes
title Near-Optimal Stochastic Bin-Packing in Large Service Systems with Time-Varying Item Sizes
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-09-22T16%3A31%3A41IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-acm_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Near-Optimal%20Stochastic%20Bin-Packing%20in%20Large%20Service%20Systems%20with%20Time-Varying%20Item%20Sizes&rft.jtitle=Performance%20evaluation%20review&rft.au=Hong,%20Yige&rft.date=2024-06-11&rft.volume=52&rft.issue=1&rft.spage=93&rft.epage=94&rft.pages=93-94&rft.issn=0163-5999&rft.eissn=1557-9484&rft_id=info:doi/10.1145/3673660.3655070&rft_dat=%3Cacm_cross%3E3655070%3C/acm_cross%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-a590-a52ebe18d22351322436f05d7bbce5a1f0761963508829ad7a22b57b76b622b33%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true