In-Memory Checkpointing for MPI Programs by XOR-Based Double-Erasure Codes
Identifieur interne : 000C77 ( Main/Merge ); précédent : 000C76; suivant : 000C78In-Memory Checkpointing for MPI Programs by XOR-Based Double-Erasure Codes
Auteurs : Gang Wang [République populaire de Chine] ; Xiaoguang Liu [République populaire de Chine] ; Ang Li [République populaire de Chine] ; Fan Zhang [République populaire de Chine]Source :
- Lecture Notes in Computer Science [ 0302-9743 ] ; 2009.
Abstract
Abstract: Today, the scale of High performance computing (HPC) systems is much larger than ever. This brings a challenge to fault tolerance of HPC systems. MPI (Message Passing Interface) is one of the most important programming tools for HPC. There are quite a few fault-tolerant extensions for MPI, such as MPICH-V, StarFish, FT-MPI and so on. Most of them are based on on-disk checkpointing. In this paper, we apply two kinds of XOR-based double-erasure codes - RDP (Row-Diagonal Parity) and B-Code to in-memory checkpointing for MPI programs. We develop scalable checkpointing/recovery algorithms which embed erasure code encoding/decoding computation into MPI collective communications operations. The experiments show that the scalable algorithms decrease communication overhead and balance computation effectively. Our approach provides highly reliable, fast in-memory checkpointing for MPI programs.
Url:
DOI: 10.1007/978-3-642-03770-2_15
Links toward previous steps (curation, corpus...)
- to stream Istex, to step Corpus: 000384
- to stream Istex, to step Curation: 000384
- to stream Istex, to step Checkpoint: 000422
Links to Exploration step
ISTEX:F6CFCB9D26FFDDB6DEB8E60F8DE484F834B85BACLe document en format XML
<record><TEI wicri:istexFullTextTei="biblStruct"><teiHeader><fileDesc><titleStmt><title xml:lang="en">In-Memory Checkpointing for MPI Programs by XOR-Based Double-Erasure Codes</title><author><name sortKey="Wang, Gang" sort="Wang, Gang" uniqKey="Wang G" first="Gang" last="Wang">Gang Wang</name></author><author><name sortKey="Liu, Xiaoguang" sort="Liu, Xiaoguang" uniqKey="Liu X" first="Xiaoguang" last="Liu">Xiaoguang Liu</name></author><author><name sortKey="Li, Ang" sort="Li, Ang" uniqKey="Li A" first="Ang" last="Li">Ang Li</name></author><author><name sortKey="Zhang, Fan" sort="Zhang, Fan" uniqKey="Zhang F" first="Fan" last="Zhang">Fan Zhang</name></author></titleStmt><publicationStmt><idno type="wicri:source">ISTEX</idno><idno type="RBID">ISTEX:F6CFCB9D26FFDDB6DEB8E60F8DE484F834B85BAC</idno><date when="2009" year="2009">2009</date><idno type="doi">10.1007/978-3-642-03770-2_15</idno><idno type="url">https://api.istex.fr/document/F6CFCB9D26FFDDB6DEB8E60F8DE484F834B85BAC/fulltext/pdf</idno><idno type="wicri:Area/Istex/Corpus">000384</idno><idno type="wicri:Area/Istex/Curation">000384</idno><idno type="wicri:Area/Istex/Checkpoint">000422</idno><idno type="wicri:doubleKey">0302-9743:2009:Wang G:in:memory:checkpointing</idno><idno type="wicri:Area/Main/Merge">000C77</idno></publicationStmt><sourceDesc><biblStruct><analytic><title level="a" type="main" xml:lang="en">In-Memory Checkpointing for MPI Programs by XOR-Based Double-Erasure Codes</title><author><name sortKey="Wang, Gang" sort="Wang, Gang" uniqKey="Wang G" first="Gang" last="Wang">Gang Wang</name><affiliation wicri:level="3"><country xml:lang="fr">République populaire de Chine</country><wicri:regionArea>Nankai-Baidu Joint Lab, College of Information Technical Science, Nankai University, 94 Weijin Road, 300071, Tianjin</wicri:regionArea><placeName><settlement type="city">Tianjin</settlement></placeName></affiliation><affiliation><wicri:noCountry code="no comma">E-mail: wgzwp@163.com</wicri:noCountry></affiliation></author><author><name sortKey="Liu, Xiaoguang" sort="Liu, Xiaoguang" uniqKey="Liu X" first="Xiaoguang" last="Liu">Xiaoguang Liu</name><affiliation wicri:level="3"><country xml:lang="fr">République populaire de Chine</country><wicri:regionArea>Nankai-Baidu Joint Lab, College of Information Technical Science, Nankai University, 94 Weijin Road, 300071, Tianjin</wicri:regionArea><placeName><settlement type="city">Tianjin</settlement></placeName></affiliation><affiliation wicri:level="1"><country wicri:rule="url">République populaire de Chine</country></affiliation></author><author><name sortKey="Li, Ang" sort="Li, Ang" uniqKey="Li A" first="Ang" last="Li">Ang Li</name><affiliation wicri:level="3"><country xml:lang="fr">République populaire de Chine</country><wicri:regionArea>Nankai-Baidu Joint Lab, College of Information Technical Science, Nankai University, 94 Weijin Road, 300071, Tianjin</wicri:regionArea><placeName><settlement type="city">Tianjin</settlement></placeName></affiliation><affiliation><wicri:noCountry code="no comma">E-mail: megathere@gmail.com</wicri:noCountry></affiliation></author><author><name sortKey="Zhang, Fan" sort="Zhang, Fan" uniqKey="Zhang F" first="Fan" last="Zhang">Fan Zhang</name><affiliation wicri:level="3"><country xml:lang="fr">République populaire de Chine</country><wicri:regionArea>Nankai-Baidu Joint Lab, College of Information Technical Science, Nankai University, 94 Weijin Road, 300071, Tianjin</wicri:regionArea><placeName><settlement type="city">Tianjin</settlement></placeName></affiliation><affiliation><wicri:noCountry code="no comma">E-mail: zhangfan555@gmail.com</wicri:noCountry></affiliation></author></analytic><monogr></monogr><series><title level="s">Lecture Notes in Computer Science</title><imprint><date>2009</date></imprint><idno type="ISSN">0302-9743</idno><idno type="eISSN">1611-3349</idno><idno type="ISSN">0302-9743</idno></series><idno type="istex">F6CFCB9D26FFDDB6DEB8E60F8DE484F834B85BAC</idno><idno type="DOI">10.1007/978-3-642-03770-2_15</idno><idno type="ChapterID">15</idno><idno type="ChapterID">Chap15</idno></biblStruct></sourceDesc><seriesStmt><idno type="ISSN">0302-9743</idno></seriesStmt></fileDesc><profileDesc><textClass></textClass><langUsage><language ident="en">en</language></langUsage></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Abstract: Today, the scale of High performance computing (HPC) systems is much larger than ever. This brings a challenge to fault tolerance of HPC systems. MPI (Message Passing Interface) is one of the most important programming tools for HPC. There are quite a few fault-tolerant extensions for MPI, such as MPICH-V, StarFish, FT-MPI and so on. Most of them are based on on-disk checkpointing. In this paper, we apply two kinds of XOR-based double-erasure codes - RDP (Row-Diagonal Parity) and B-Code to in-memory checkpointing for MPI programs. We develop scalable checkpointing/recovery algorithms which embed erasure code encoding/decoding computation into MPI collective communications operations. The experiments show that the scalable algorithms decrease communication overhead and balance computation effectively. Our approach provides highly reliable, fast in-memory checkpointing for MPI programs.</div></front></TEI></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Ticri/CIDE/explor/CyberinfraV1/Data/Main/Merge
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000C77 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Merge/biblio.hfd -nk 000C77 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Ticri/CIDE
|area= CyberinfraV1
|flux= Main
|étape= Merge
|type= RBID
|clé= ISTEX:F6CFCB9D26FFDDB6DEB8E60F8DE484F834B85BAC
|texte= In-Memory Checkpointing for MPI Programs by XOR-Based Double-Erasure Codes
}}
| This area was generated with Dilib version V0.6.25. |  |