Lattice Quantum Chromodynamics (QCD) is one of the most challenging applications running on massively parallel supercomputers. To reproduce these physical phenomena on a supercomputer, a precise simulation is demanded requiring well optimized and scalable code. We have optimized lattice QCD programs on Blue Gene family supercomputers and shown the strength in lattice QCD simulation. Here we optimized on the third generation Blue Gene/Q supercomputer; i) by changing the data layout, ii) by exploiting new SIMD instruction sets, and iii) by pipelining boundary data exchange to overlap communication and calculation. The optimized lattice QCD program shows excellent weak scalability on the large scale Blue Gene/Q system, and with 16 racks we sustained 1.08 Pflop/s, 32.1% of the theoretical peak performance, including the conjugate gradient solver routines.
{"title":"Peta-scale Lattice Quantum Chromodynamics on a Blue Gene/Q supercomputer","authors":"J. Doi","doi":"10.1109/SC.2012.96","DOIUrl":"https://doi.org/10.1109/SC.2012.96","url":null,"abstract":"Lattice Quantum Chromodynamics (QCD) is one of the most challenging applications running on massively parallel supercomputers. To reproduce these physical phenomena on a supercomputer, a precise simulation is demanded requiring well optimized and scalable code. We have optimized lattice QCD programs on Blue Gene family supercomputers and shown the strength in lattice QCD simulation. Here we optimized on the third generation Blue Gene/Q supercomputer; i) by changing the data layout, ii) by exploiting new SIMD instruction sets, and iii) by pipelining boundary data exchange to overlap communication and calculation. The optimized lattice QCD program shows excellent weak scalability on the large scale Blue Gene/Q system, and with 16 racks we sustained 1.08 Pflop/s, 32.1% of the theoretical peak performance, including the conjugate gradient solver routines.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124092224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Bayesian networks (BN) are probabilistic graphical models which are widely utilized in various research areas, including modeling complex biological interactions in the cell. Learning the structure of a BN is an NP-hard problem and exact solutions are limited to a few tens of variables. In this work, we present a parallel BN structure learning algorithm that combines principles of both heuristic and exact approaches and facilitates learning of larger networks. We demonstrate the applicability of our approach by an implementation on a Cray AMD cluster, and present experimental results for the problem of inferring gene networks. Our approach is work-optimal and achieves nearly perfect scaling.
{"title":"Parallel Bayesian network structure learning with application to gene networks","authors":"Olga Nikolova, S. Aluru","doi":"10.1109/SC.2012.90","DOIUrl":"https://doi.org/10.1109/SC.2012.90","url":null,"abstract":"Bayesian networks (BN) are probabilistic graphical models which are widely utilized in various research areas, including modeling complex biological interactions in the cell. Learning the structure of a BN is an NP-hard problem and exact solutions are limited to a few tens of variables. In this work, we present a parallel BN structure learning algorithm that combines principles of both heuristic and exact approaches and facilitates learning of larger networks. We demonstrate the applicability of our approach by an implementation on a Cray AMD cluster, and present experimental results for the problem of inferring gene networks. Our approach is work-optimal and achieves nearly perfect scaling.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130176237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
T. Bui-Thanh, Carsten Burstedde, O. Ghattas, James Martin, G. Stadler, L. Wilcox
Quantifying uncertainties in large-scale simulations has emerged as the central challenge facing CS&E. When the simulations require supercomputers, and uncertain parameter dimensions are large, conventional UQ methods fail. Here we address uncertainty quantification for large-scale inverse problems in a Bayesian inference framework: given data and model uncertainties, find the pdf describing parameter uncertainties. To overcome the curse of dimensionality of conventional methods, we exploit the fact that the data are typically informative about low-dimensional manifolds of parameter space to construct low rank approximations of the covariance matrix of the posterior pdf via a matrix-free randomized method. We obtain a method that scales independently of the forward problem dimension, the uncertain parameter dimension, the data dimension, and the number of cores. We apply the method to the Bayesian solution of an inverse problem in 3D global seismic wave propagation with over one million uncertain earth model parameters, 630 million wave propagation unknowns, on up to 262K cores, for which we obtain a factor of over 2000 reduction in problem dimension. This makes UQ tractable for the inverse problem.
{"title":"Extreme-scale UQ for Bayesian inverse problems governed by PDEs","authors":"T. Bui-Thanh, Carsten Burstedde, O. Ghattas, James Martin, G. Stadler, L. Wilcox","doi":"10.1109/SC.2012.56","DOIUrl":"https://doi.org/10.1109/SC.2012.56","url":null,"abstract":"Quantifying uncertainties in large-scale simulations has emerged as the central challenge facing CS&E. When the simulations require supercomputers, and uncertain parameter dimensions are large, conventional UQ methods fail. Here we address uncertainty quantification for large-scale inverse problems in a Bayesian inference framework: given data and model uncertainties, find the pdf describing parameter uncertainties. To overcome the curse of dimensionality of conventional methods, we exploit the fact that the data are typically informative about low-dimensional manifolds of parameter space to construct low rank approximations of the covariance matrix of the posterior pdf via a matrix-free randomized method. We obtain a method that scales independently of the forward problem dimension, the uncertain parameter dimension, the data dimension, and the number of cores. We apply the method to the Bayesian solution of an inverse problem in 3D global seismic wave propagation with over one million uncertain earth model parameters, 630 million wave propagation unknowns, on up to 262K cores, for which we obtain a factor of over 2000 reduction in problem dimension. This makes UQ tractable for the inverse problem.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120964808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Vignesh T. Ravi, M. Becchi, G. Agrawal, S. Chakradhar
Heterogeneous computing nodes are becoming commonplace today, and recent trends strongly indicate that clusters, supercomputers, and cloud environments will increasingly host more heterogeneous resources, with some being massively parallel (e.g., GPU). With such heterogeneous environments becoming common, it is important to revisit scheduling problems for clusters and cloud environments. In this paper, we formulate and address the problem of value-driven scheduling of independent jobs on heterogeneous clusters, which captures both the urgency and relative priority of jobs. Our overall scheduling goal is to maximize the aggregate value or yield of all jobs. Exploiting the portability available from the underlying programming model, we propose four novel scheduling schemes that can automatically and dynamically map jobs onto heterogeneous resources. Additionally, to improve the utilization of massively parallel resources, we also propose heuristics to automatically decide when and which jobs can share a single resource.
{"title":"ValuePack: Value-based scheduling framework for CPU-GPU clusters","authors":"Vignesh T. Ravi, M. Becchi, G. Agrawal, S. Chakradhar","doi":"10.1109/SC.2012.111","DOIUrl":"https://doi.org/10.1109/SC.2012.111","url":null,"abstract":"Heterogeneous computing nodes are becoming commonplace today, and recent trends strongly indicate that clusters, supercomputers, and cloud environments will increasingly host more heterogeneous resources, with some being massively parallel (e.g., GPU). With such heterogeneous environments becoming common, it is important to revisit scheduling problems for clusters and cloud environments. In this paper, we formulate and address the problem of value-driven scheduling of independent jobs on heterogeneous clusters, which captures both the urgency and relative priority of jobs. Our overall scheduling goal is to maximize the aggregate value or yield of all jobs. Exploiting the portability available from the underlying programming model, we propose four novel scheduling schemes that can automatically and dynamically map jobs onto heterogeneous resources. Additionally, to improve the utilization of massively parallel resources, we also propose heuristics to automatically decide when and which jobs can share a single resource.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115608322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Many scientific applications operate in a bulk-synchronous mode of iterative communication and computation steps. Even though the communication steps happen at the same logical time, important patterns such as stencil computations cannot be expressed as collective communications in MPI. We demonstrate how neighborhood collective operations allow to specify arbitrary collective communication relations during runtime and enable optimizations similar to traditional collective calls. We show a number of optimization opportunities and algorithms for different communication scenarios. We also show how users can assert constraints that provide additional optimization opportunities in a portable way. We demonstrate the utility of all described optimizations in a highly optimized implementation of neighborhood collective operations. Our communication and protocol optimizations result in a performance improvement of up to a factor of two for small stencil communications. We found that, for some patterns, our optimization heuristics automatically generate communication schedules that are comparable to hand-tuned collectives. With those optimizations in place, we are able to accelerate arbitrary collective communication patterns, such as regular and irregular stencils with optimization methods for collective communications. We expect that our methods will influence the design of future MPI libraries and provide a significant performance benefit on large-scale systems.
{"title":"Optimization principles for collective neighborhood communications","authors":"T. Hoefler, Timo Schneider","doi":"10.1109/SC.2012.86","DOIUrl":"https://doi.org/10.1109/SC.2012.86","url":null,"abstract":"Many scientific applications operate in a bulk-synchronous mode of iterative communication and computation steps. Even though the communication steps happen at the same logical time, important patterns such as stencil computations cannot be expressed as collective communications in MPI. We demonstrate how neighborhood collective operations allow to specify arbitrary collective communication relations during runtime and enable optimizations similar to traditional collective calls. We show a number of optimization opportunities and algorithms for different communication scenarios. We also show how users can assert constraints that provide additional optimization opportunities in a portable way. We demonstrate the utility of all described optimizations in a highly optimized implementation of neighborhood collective operations. Our communication and protocol optimizations result in a performance improvement of up to a factor of two for small stencil communications. We found that, for some patterns, our optimization heuristics automatically generate communication schedules that are comparable to hand-tuned collectives. With those optimizations in place, we are able to accelerate arbitrary collective communication patterns, such as regular and irregular stencils with optimization methods for collective communications. We expect that our methods will influence the design of future MPI libraries and provide a significant performance benefit on large-scale systems.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127064742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Janine Bennett, H. Abbasi, P. Bremer, R. Grout, A. Gyulassy, Tong Jin, S. Klasky, H. Kolla, M. Parashar, Valerio Pascucci, P. Pébay, D. Thompson, Hongfeng Yu, Fan Zhang, Jacqueline H. Chen
With the onset of extreme-scale computing, I/O constraints make it increasingly difficult for scientists to save a sufficient amount of raw simulation data to persistent storage. One potential solution is to change the data analysis pipeline from a post-process centric to a concurrent approach based on either in-situ or in-transit processing. In this context computations are considered in-situ if they utilize the primary compute resources, while in-transit processing refers to offloading computations to a set of secondary resources using asynchronous data transfers. In this paper we explore the design and implementation of three common analysis techniques typically performed on large-scale scientific simulations: topological analysis, descriptive statistics, and visualization. We summarize algorithmic developments, describe a resource scheduling system to coordinate the execution of various analysis workflows, and discuss our implementation using the DataSpaces and ADIOS frameworks that support efficient data movement between in-situ and in-transit computations. We demonstrate the efficiency of our lightweight, flexible framework by deploying it on the Jaguar XK6 to analyze data generated by S3D, a massively parallel turbulent combustion code. Our framework allows scientists dealing with the data deluge at extreme scale to perform analyses at increased temporal resolutions, mitigate I/O costs, and significantly improve the time to insight.
{"title":"Combining in-situ and in-transit processing to enable extreme-scale scientific analysis","authors":"Janine Bennett, H. Abbasi, P. Bremer, R. Grout, A. Gyulassy, Tong Jin, S. Klasky, H. Kolla, M. Parashar, Valerio Pascucci, P. Pébay, D. Thompson, Hongfeng Yu, Fan Zhang, Jacqueline H. Chen","doi":"10.1109/SC.2012.31","DOIUrl":"https://doi.org/10.1109/SC.2012.31","url":null,"abstract":"With the onset of extreme-scale computing, I/O constraints make it increasingly difficult for scientists to save a sufficient amount of raw simulation data to persistent storage. One potential solution is to change the data analysis pipeline from a post-process centric to a concurrent approach based on either in-situ or in-transit processing. In this context computations are considered in-situ if they utilize the primary compute resources, while in-transit processing refers to offloading computations to a set of secondary resources using asynchronous data transfers. In this paper we explore the design and implementation of three common analysis techniques typically performed on large-scale scientific simulations: topological analysis, descriptive statistics, and visualization. We summarize algorithmic developments, describe a resource scheduling system to coordinate the execution of various analysis workflows, and discuss our implementation using the DataSpaces and ADIOS frameworks that support efficient data movement between in-situ and in-transit computations. We demonstrate the efficiency of our lightweight, flexible framework by deploying it on the Jaguar XK6 to analyze data generated by S3D, a massively parallel turbulent combustion code. Our framework allows scientists dealing with the data deluge at extreme scale to perform analyses at increased temporal resolutions, mitigate I/O costs, and significantly improve the time to insight.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123692912","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Greg Faanes, A. Bataineh, D. Roweth, T. Court, E. Froese, Robert Alverson, Tim Johnson, Joe Kopnick, Mike Higgins, James Reinhard
Higher global bandwidth requirement for many applications and lower network cost have motivated the use of the Dragonfly network topology for high performance computing systems. In this paper we present the architecture of the Cray Cascade system, a distributed memory system based on the Dragonfly [1] network topology. We describe the structure of the system, its Dragonfly network and the routing algorithms. We describe a set of advanced features supporting both mainstream high performance computing applications and emerging global address space programing models. We present a combination of performance results from prototype systems and simulation data for large systems. We demonstrate the value of the Dragonfly topology and the benefits obtained through extensive use of adaptive routing.
{"title":"Cray Cascade: A scalable HPC system based on a Dragonfly network","authors":"Greg Faanes, A. Bataineh, D. Roweth, T. Court, E. Froese, Robert Alverson, Tim Johnson, Joe Kopnick, Mike Higgins, James Reinhard","doi":"10.1109/SC.2012.39","DOIUrl":"https://doi.org/10.1109/SC.2012.39","url":null,"abstract":"Higher global bandwidth requirement for many applications and lower network cost have motivated the use of the Dragonfly network topology for high performance computing systems. In this paper we present the architecture of the Cray Cascade system, a distributed memory system based on the Dragonfly [1] network topology. We describe the structure of the system, its Dragonfly network and the routing algorithms. We describe a set of advanced features supporting both mainstream high performance computing applications and emerging global address space programing models. We present a combination of performance results from prototype systems and simulation data for large systems. We demonstrate the value of the Dragonfly topology and the benefits obtained through extensive use of adaptive routing.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124535842","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Stephen L. Olivier, B. Supinski, M. Schulz, J. Prins
Task parallelism raises the level of abstraction in shared memory parallel programming to simplify the development of complex applications. However, task parallel applications can exhibit poor performance due to thread idleness, scheduling overheads, and work time inflation -- additional time spent by threads in a multithreaded computation beyond the time required to perform the same work in a sequential computation. We identify the contributions of each factor to lost efficiency in various task parallel OpenMP applications and diagnose the causes of work time inflation in those applications. Increased data access latency can cause significant work time inflation in NUMA systems. Our locality framework for task parallel OpenMP programs mitigates this cause of work time inflation. Our extensions to the Qthreads library demonstrate that locality-aware scheduling can improve performance up to 3X compared to the Intel OpenMP task scheduler.
{"title":"Characterizing and mitigating work time inflation in task parallel programs","authors":"Stephen L. Olivier, B. Supinski, M. Schulz, J. Prins","doi":"10.3233/SPR-130369","DOIUrl":"https://doi.org/10.3233/SPR-130369","url":null,"abstract":"Task parallelism raises the level of abstraction in shared memory parallel programming to simplify the development of complex applications. However, task parallel applications can exhibit poor performance due to thread idleness, scheduling overheads, and work time inflation -- additional time spent by threads in a multithreaded computation beyond the time required to perform the same work in a sequential computation. We identify the contributions of each factor to lost efficiency in various task parallel OpenMP applications and diagnose the causes of work time inflation in those applications. Increased data access latency can cause significant work time inflation in NUMA systems. Our locality framework for task parallel OpenMP programs mitigates this cause of work time inflation. Our extensions to the Qthreads library demonstrate that locality-aware scheduling can improve performance up to 3X compared to the Intel OpenMP task scheduler.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123979066","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dirk Meister, J. Kaiser, A. Brinkmann, Toni Cortes, Michael Kuhn, J. Kunkel
Deduplication is a storage saving technique that is highly successful in enterprise backup environments. On a file system, a single data block might be stored multiple times across different files, for example, multiple versions of a file might exist that are mostly identical. With deduplication, this data replication is localized and redundancy is removed -- by storing data just once, all files that use identical regions refer to the same unique data. The most common approach splits file data into chunks and calculates a cryptographic fingerprint for each chunk. By checking if the fingerprint has already been stored, a chunk is classified as redundant or unique. Only unique chunks are stored. This paper presents the first study on the potential of data deduplication in HPC centers, which belong to the most demanding storage producers. We have quantitatively assessed this potential for capacity reduction for 4 data centers (BSC, DKRZ, RENCI, RWTH). In contrast to previous deduplication studies focusing mostly on backup data, we have analyzed over one PB (1212 TB) of online file system data. The evaluation shows that typically 20% to 30% of this online data can be removed by applying data deduplication techniques, peaking up to 70% for some data sets. This reduction can only be achieved by a subfile deduplication approach, while approaches based on whole-file comparisons only lead to small capacity savings.
{"title":"A study on data deduplication in HPC storage systems","authors":"Dirk Meister, J. Kaiser, A. Brinkmann, Toni Cortes, Michael Kuhn, J. Kunkel","doi":"10.1109/SC.2012.14","DOIUrl":"https://doi.org/10.1109/SC.2012.14","url":null,"abstract":"Deduplication is a storage saving technique that is highly successful in enterprise backup environments. On a file system, a single data block might be stored multiple times across different files, for example, multiple versions of a file might exist that are mostly identical. With deduplication, this data replication is localized and redundancy is removed -- by storing data just once, all files that use identical regions refer to the same unique data. The most common approach splits file data into chunks and calculates a cryptographic fingerprint for each chunk. By checking if the fingerprint has already been stored, a chunk is classified as redundant or unique. Only unique chunks are stored. This paper presents the first study on the potential of data deduplication in HPC centers, which belong to the most demanding storage producers. We have quantitatively assessed this potential for capacity reduction for 4 data centers (BSC, DKRZ, RENCI, RWTH). In contrast to previous deduplication studies focusing mostly on backup data, we have analyzed over one PB (1212 TB) of online file system data. The evaluation shows that typically 20% to 30% of this online data can be removed by applying data deduplication techniques, peaking up to 70% for some data sets. This reduction can only be achieved by a subfile deduplication approach, while approaches based on whole-file comparisons only lead to small capacity savings.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125521184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Riesen, Kurt B. Ferreira, D. D. Silva, Pierre Lemarinier, D. Arnold, P. Bridges
Current fault tolerance protocols are not sufficiently scalable for the exascale era. The most-widely used method, coordinated checkpointing, places enormous demands on the I/O subsystem and imposes frequent synchronizations. Uncoordinated protocols use message logging which introduces message rate limitations or undesired memory and storage requirements to hold payload and event logs. In this paper we propose a combination of several techniques, namely coordinated checkpointing, optimistic message logging, and a protocol that glues them together. This combination eliminates some of the drawbacks of each individual approach and proves to be an alternative for many types of exascale applications. We evaluate performance and scaling characteristics of this combination using simulation and a partial implementation. While not a universal solution, the combined protocol is suitable for a large range of existing and future applications that use coordinated checkpointing and enhances their scalability.
{"title":"Alleviating scalability issues of checkpointing protocols","authors":"R. Riesen, Kurt B. Ferreira, D. D. Silva, Pierre Lemarinier, D. Arnold, P. Bridges","doi":"10.1109/SC.2012.18","DOIUrl":"https://doi.org/10.1109/SC.2012.18","url":null,"abstract":"Current fault tolerance protocols are not sufficiently scalable for the exascale era. The most-widely used method, coordinated checkpointing, places enormous demands on the I/O subsystem and imposes frequent synchronizations. Uncoordinated protocols use message logging which introduces message rate limitations or undesired memory and storage requirements to hold payload and event logs. In this paper we propose a combination of several techniques, namely coordinated checkpointing, optimistic message logging, and a protocol that glues them together. This combination eliminates some of the drawbacks of each individual approach and proves to be an alternative for many types of exascale applications. We evaluate performance and scaling characteristics of this combination using simulation and a partial implementation. While not a universal solution, the combined protocol is suitable for a large range of existing and future applications that use coordinated checkpointing and enhances their scalability.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"135 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127593460","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：