Workload characterization of input/output intensive parallel applications

Workload characterization of input/output intensive parallel applications

The broadening disparity in the performance of input/output (I/O) devices and the performance of processors and communication links on parallel systems is a major obstacle to achieving high performance for a wide range of parallel applications. I/O hardware and file system parallelism are the keys t...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Smirni, Evgenia, Reed, Daniel A.
Format: Buchkapitel
Sprache:eng
Schlagworte:
Access Pattern
Application Programming Interface
Execution Time
File System
Parallel Application
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 180
container_issue
container_start_page 169
container_title
container_volume
creator Smirni, Evgenia
Reed, Daniel A.
description The broadening disparity in the performance of input/output (I/O) devices and the performance of processors and communication links on parallel systems is a major obstacle to achieving high performance for a wide range of parallel applications. I/O hardware and file system parallelism are the keys to bridging this performance gap. A prerequisite to the development of efficient parallel file systems is detailed characterization of the I/O demands of parallel applications. In this paper, we present a comparative study of the I/O access patterns commonly found in I/O intensive parallel applications. Using the Pablo performance analysis environment and its I/O extensions we captured application I/O access patterns and analyzed their interactions with current parallel I/O systems. This analysis has proven instrumental in guiding the development of new application programming interfaces (APIs) for parallel file systems and in developing effective file system policies that can adaptively respond to complex application I/O requirements.
doi_str_mv 10.1007/BFb0022205
format Book Chapter
fullrecord <record><control><sourceid>springer</sourceid><recordid>TN_cdi_springer_books_10_1007_BFb0022205</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>springer_books_10_1007_BFb0022205</sourcerecordid><originalsourceid>FETCH-LOGICAL-c1445-8abcdd9beedeebd371b286bc7559745fa35e27b9a9bb432d20326f948172024c3</originalsourceid><addsrcrecordid>eNpFkD9PwzAUxM0_ibR04RNkZAl9z8-O4xGqFipVYgExRrbjQGgUR3HKwKcnFCSm0-l-Op2OsWuEWwRQy_uNBeCcgzxhM5ICco2E-SlLMEfMiIQ-YwutimNGCKjPWQIEPNNK0CWbxfgBU4XSPGHb1zDs22Cq1L2bwbjRD82XGZvQpaFOm64_jMtwGCeZzOi72Hz6tJ_ItvVtavq-bdwRj1fsojZt9Is_nbOXzfp59Zjtnh62q7td5lAImRXGuqrS1vvKe1uRQsuL3Dol5TRP1oak58pqo60VxCsOxPNaiwIVBy4czdnNb2_sh6Z780NpQ9jHEqH8-af8_4e-AQ-nVGY</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>book_chapter</recordtype></control><display><type>book_chapter</type><title>Workload characterization of input/output intensive parallel applications</title><source>Springer Books</source><creator>Smirni, Evgenia ; Reed, Daniel A.</creator><contributor>Rubino, Gerardo ; Plateau, Brigitte ; Marie, Raymond ; Calzarossa, Maria</contributor><creatorcontrib>Smirni, Evgenia ; Reed, Daniel A. ; Rubino, Gerardo ; Plateau, Brigitte ; Marie, Raymond ; Calzarossa, Maria</creatorcontrib><description>The broadening disparity in the performance of input/output (I/O) devices and the performance of processors and communication links on parallel systems is a major obstacle to achieving high performance for a wide range of parallel applications. I/O hardware and file system parallelism are the keys to bridging this performance gap. A prerequisite to the development of efficient parallel file systems is detailed characterization of the I/O demands of parallel applications. In this paper, we present a comparative study of the I/O access patterns commonly found in I/O intensive parallel applications. Using the Pablo performance analysis environment and its I/O extensions we captured application I/O access patterns and analyzed their interactions with current parallel I/O systems. This analysis has proven instrumental in guiding the development of new application programming interfaces (APIs) for parallel file systems and in developing effective file system policies that can adaptively respond to complex application I/O requirements.</description><identifier>ISSN: 0302-9743</identifier><identifier>ISBN: 9783540631019</identifier><identifier>ISBN: 3540631011</identifier><identifier>EISSN: 1611-3349</identifier><identifier>EISBN: 3540691316</identifier><identifier>EISBN: 9783540691310</identifier><identifier>DOI: 10.1007/BFb0022205</identifier><language>eng</language><publisher>Berlin, Heidelberg: Springer Berlin Heidelberg</publisher><subject>Access Pattern ; Application Programming Interface ; Execution Time ; File System ; Parallel Application</subject><ispartof>Computer Performance Evaluation Modelling Techniques and Tools, 2005, p.169-180</ispartof><rights>Springer-Verlag Berlin Heidelberg 1997</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c1445-8abcdd9beedeebd371b286bc7559745fa35e27b9a9bb432d20326f948172024c3</citedby><relation>Lecture Notes in Computer Science</relation></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://link.springer.com/content/pdf/10.1007/BFb0022205$$EPDF$$P50$$Gspringer$$H</linktopdf><linktohtml>$$Uhttps://link.springer.com/10.1007/BFb0022205$$EHTML$$P50$$Gspringer$$H</linktohtml><link.rule.ids>776,777,781,790,27906,38236,41423,42492</link.rule.ids></links><search><contributor>Rubino, Gerardo</contributor><contributor>Plateau, Brigitte</contributor><contributor>Marie, Raymond</contributor><contributor>Calzarossa, Maria</contributor><creatorcontrib>Smirni, Evgenia</creatorcontrib><creatorcontrib>Reed, Daniel A.</creatorcontrib><title>Workload characterization of input/output intensive parallel applications</title><title>Computer Performance Evaluation Modelling Techniques and Tools</title><description>The broadening disparity in the performance of input/output (I/O) devices and the performance of processors and communication links on parallel systems is a major obstacle to achieving high performance for a wide range of parallel applications. I/O hardware and file system parallelism are the keys to bridging this performance gap. A prerequisite to the development of efficient parallel file systems is detailed characterization of the I/O demands of parallel applications. In this paper, we present a comparative study of the I/O access patterns commonly found in I/O intensive parallel applications. Using the Pablo performance analysis environment and its I/O extensions we captured application I/O access patterns and analyzed their interactions with current parallel I/O systems. This analysis has proven instrumental in guiding the development of new application programming interfaces (APIs) for parallel file systems and in developing effective file system policies that can adaptively respond to complex application I/O requirements.</description><subject>Access Pattern</subject><subject>Application Programming Interface</subject><subject>Execution Time</subject><subject>File System</subject><subject>Parallel Application</subject><issn>0302-9743</issn><issn>1611-3349</issn><isbn>9783540631019</isbn><isbn>3540631011</isbn><isbn>3540691316</isbn><isbn>9783540691310</isbn><fulltext>true</fulltext><rsrctype>book_chapter</rsrctype><creationdate>2005</creationdate><recordtype>book_chapter</recordtype><sourceid/><recordid>eNpFkD9PwzAUxM0_ibR04RNkZAl9z8-O4xGqFipVYgExRrbjQGgUR3HKwKcnFCSm0-l-Op2OsWuEWwRQy_uNBeCcgzxhM5ICco2E-SlLMEfMiIQ-YwutimNGCKjPWQIEPNNK0CWbxfgBU4XSPGHb1zDs22Cq1L2bwbjRD82XGZvQpaFOm64_jMtwGCeZzOi72Hz6tJ_ItvVtavq-bdwRj1fsojZt9Is_nbOXzfp59Zjtnh62q7td5lAImRXGuqrS1vvKe1uRQsuL3Dol5TRP1oak58pqo60VxCsOxPNaiwIVBy4czdnNb2_sh6Z780NpQ9jHEqH8-af8_4e-AQ-nVGY</recordid><startdate>20050616</startdate><enddate>20050616</enddate><creator>Smirni, Evgenia</creator><creator>Reed, Daniel A.</creator><general>Springer Berlin Heidelberg</general><scope/></search><sort><creationdate>20050616</creationdate><title>Workload characterization of input/output intensive parallel applications</title><author>Smirni, Evgenia ; Reed, Daniel A.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c1445-8abcdd9beedeebd371b286bc7559745fa35e27b9a9bb432d20326f948172024c3</frbrgroupid><rsrctype>book_chapters</rsrctype><prefilter>book_chapters</prefilter><language>eng</language><creationdate>2005</creationdate><topic>Access Pattern</topic><topic>Application Programming Interface</topic><topic>Execution Time</topic><topic>File System</topic><topic>Parallel Application</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Smirni, Evgenia</creatorcontrib><creatorcontrib>Reed, Daniel A.</creatorcontrib></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Smirni, Evgenia</au><au>Reed, Daniel A.</au><au>Rubino, Gerardo</au><au>Plateau, Brigitte</au><au>Marie, Raymond</au><au>Calzarossa, Maria</au><format>book</format><genre>bookitem</genre><ristype>CHAP</ristype><atitle>Workload characterization of input/output intensive parallel applications</atitle><btitle>Computer Performance Evaluation Modelling Techniques and Tools</btitle><seriestitle>Lecture Notes in Computer Science</seriestitle><date>2005-06-16</date><risdate>2005</risdate><spage>169</spage><epage>180</epage><pages>169-180</pages><issn>0302-9743</issn><eissn>1611-3349</eissn><isbn>9783540631019</isbn><isbn>3540631011</isbn><eisbn>3540691316</eisbn><eisbn>9783540691310</eisbn><abstract>The broadening disparity in the performance of input/output (I/O) devices and the performance of processors and communication links on parallel systems is a major obstacle to achieving high performance for a wide range of parallel applications. I/O hardware and file system parallelism are the keys to bridging this performance gap. A prerequisite to the development of efficient parallel file systems is detailed characterization of the I/O demands of parallel applications. In this paper, we present a comparative study of the I/O access patterns commonly found in I/O intensive parallel applications. Using the Pablo performance analysis environment and its I/O extensions we captured application I/O access patterns and analyzed their interactions with current parallel I/O systems. This analysis has proven instrumental in guiding the development of new application programming interfaces (APIs) for parallel file systems and in developing effective file system policies that can adaptively respond to complex application I/O requirements.</abstract><cop>Berlin, Heidelberg</cop><pub>Springer Berlin Heidelberg</pub><doi>10.1007/BFb0022205</doi><tpages>12</tpages></addata></record>
fulltext fulltext
identifier ISSN: 0302-9743
ispartof Computer Performance Evaluation Modelling Techniques and Tools, 2005, p.169-180
issn 0302-9743
1611-3349
language eng
recordid cdi_springer_books_10_1007_BFb0022205
source Springer Books
subjects Access Pattern
Application Programming Interface
Execution Time
File System
Parallel Application
title Workload characterization of input/output intensive parallel applications
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-20T22%3A01%3A59IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-springer&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=bookitem&rft.atitle=Workload%20characterization%20of%20input/output%20intensive%20parallel%20applications&rft.btitle=Computer%20Performance%20Evaluation%20Modelling%20Techniques%20and%20Tools&rft.au=Smirni,%20Evgenia&rft.date=2005-06-16&rft.spage=169&rft.epage=180&rft.pages=169-180&rft.issn=0302-9743&rft.eissn=1611-3349&rft.isbn=9783540631019&rft.isbn_list=3540631011&rft_id=info:doi/10.1007/BFb0022205&rft_dat=%3Cspringer%3Espringer_books_10_1007_BFb0022205%3C/springer%3E%3Curl%3E%3C/url%3E&rft.eisbn=3540691316&rft.eisbn_list=9783540691310&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true