2005 IEEE International Conference on Cluster Computing最新文献_第2页

A Cost/Benefit Estimating Service for Mapping Parallel Applications on Heterogeneous Clusters 异构集群上并行应用映射的成本/收益评估服务

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347062

D. Katramatos, S. Chapin

{"title":"A Cost/Benefit Estimating Service for Mapping Parallel Applications on Heterogeneous Clusters","authors":"D. Katramatos, S. Chapin","doi":"10.1109/CLUSTR.2005.347062","DOIUrl":"https://doi.org/10.1109/CLUSTR.2005.347062","url":null,"abstract":"Matching the resource requirements of a parallel application to the available resources of a large, heterogeneous cluster is a key requirement in effectively scheduling the application tasks on the nodes of the cluster. This paper describes the cost/benefit estimating service (CBES), a runtime scheduling system targeted at finding highly effective schedules (or mappings) of tasks on nodes. CBES relies on its own infrastructure to gather and maintain static and dynamic information profiles for the computing system and the applications of interest. At the core of CBES is a mapping evaluation module which evaluates candidate application mappings on the basis of shortest execution times. By default, CBES uses a simulated-annealing based scheduler to select mappings. The paper presents the design, initial implementation, and test results of CBES on the Centurion cluster at the University of Virginia and the Orange Grove cluster at Syracuse University. These tests demonstrated that the exploitation of internode communication speed differences due to network heterogeneity can yield speedups of over 10% between same architecture nodes. The maximum observed speedup across architectures for the best vs. worst mapping scenarios of the same application was over 36%, while the corresponding average case speedup was approximately 30%","PeriodicalId":255312,"journal":{"name":"2005 IEEE International Conference on Cluster Computing","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117187701","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}

引用次数: 7

Experiment Management and Analysis with perfbase 基于perfbase的实验管理与分析

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347052

Joachim Worringen

引用次数: 10

A pipelined data-parallel algorithm for ILP 一种用于ILP的流水线数据并行算法

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347059

N.A. Fonseca, Fernando M A Silva, V. S. Costa, Rui Camacho

引用次数: 12

Performance Effects of Interrupt Throttle Rate on Linux Clusters using Intel Gigabit Network Adapters 使用Intel千兆网卡的Linux集群中断节流率对性能的影响

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347089

Baris Guler, R. Radhakrishnan, Ronald Pepper

{"title":"Performance Effects of Interrupt Throttle Rate on Linux Clusters using Intel Gigabit Network Adapters","authors":"Baris Guler, R. Radhakrishnan, Ronald Pepper","doi":"10.1109/CLUSTR.2005.347089","DOIUrl":"https://doi.org/10.1109/CLUSTR.2005.347089","url":null,"abstract":"Summary form only given. Many high performance computing clusters (HPCC) are still built using gigabit Ethernet as the interconnect connecting all the computing nodes even though there are faster (lower latency and higher bandwidth) alternatives such as Infiniband and Myrinet. The choice of interconnect mainly depends on the parallel application communication characteristics as well as budget requirements since the faster alternatives are much more expensive compared to gigabit Ethernet especially at lower node counts. Some applications require lower latency interconnect since they communicate more frequently but send relatively small messages, and others can be sending infrequent but large messages thus requiring a higher bandwidth interconnect. Since PCs, workstations and servers are designed for server-client type of environment, network interface card (NIC) drivers are usually optimized for specific network traffic patterns by using several interrupt moderation techniques/parameters, specifically interrupt throttle rate (ITR). Since in an HPCC environment the parallel application communication characteristics (i.e. network traffic patterns) are usually different than the default setting, an ITR value has to be identified to achieve best overall system performance for each type of application. This poster will present the case for why this is an important area in high-performance computing clusters connected using gigabit interconnects. It will present methodologies to tune the interrupt throttle rate parameter given to the driver to achieve a balance between application and network performance. Performance results on typical applications will be shown on different clusters","PeriodicalId":255312,"journal":{"name":"2005 IEEE International Conference on Cluster Computing","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114698949","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}

引用次数: 0

A New Allocation Scheme for Parallel Applications with Deadline and Security Constraints on Clusters 一种簇上有时限和安全约束的并行应用分配新方案

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347057

T. Xie, X. Qin

引用次数: 30

Memory Management Support for Multi-Programmed Remote Direct Memory Access (RDMA) Systems 多程序远程直接内存访问(RDMA)系统的内存管理支持

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347031

K. Magoutis

{"title":"Memory Management Support for Multi-Programmed Remote Direct Memory Access (RDMA) Systems","authors":"K. Magoutis","doi":"10.1109/CLUSTR.2005.347031","DOIUrl":"https://doi.org/10.1109/CLUSTR.2005.347031","url":null,"abstract":"Current operating systems offer basic support for network interface controllers (NICs) supporting remote direct memory access (RDMA). Such support typically consists of a device driver responsible for configuring communication channels between the device and user-level processes but not involved in data transfer. Unlike standard NICs, RDMA-capable devices incorporate significant memory resources for address translation purposes. In a multi-programmed operating system (OS) environment, these memory resources must be efficiently shareable by multiple processes. For such sharing to occur in a fair manner, the OS and the device must cooperate to arbitrate access to NIC memory, similar to the way CPUs and OSes cooperate to arbitrate access to translation lookaside buffers (TLBs) or physical memory. A problem with this approach is that today's RDMA NICs are not integrated into the functions provided by OS memory management systems. As a result, RDMA NIC hardware resources are often monopolized by a single application. In this paper, I propose two practical mechanisms to address this problem: (a) Use of RDMA only in kernel-resident I/O subsystems, transparent to user-level software; (b) An extended registration API and a kernel upcall mechanism delivering NIC TLB entry replacement notifications to user-level libraries. Both options are designed to re-instate the multiprogramming principles that are violated in early commercial RDMA systems","PeriodicalId":255312,"journal":{"name":"2005 IEEE International Conference on Cluster Computing","volume":"192 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133948035","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}

引用次数: 6

Processes Distribution of Homogeneous Parallel Linear Algebra Routines on Heterogeneous Clusters 异质簇上齐次平行线性代数例程的过程分布

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347021

J. Cuenca, Luis-Pedro García, D. Giménez, J. Dongarra

引用次数: 15

Toward an Optimal Redundancy Strategy for Distributed Computations 面向分布式计算的最优冗余策略

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-27 DOI: 10.1109/CLUSTR.2005.347045

Doug Szajda, Barry Lawson, Jason Owen

引用次数: 22

Topology-Based Meshes for Local Communication in Data-Parallel Applications 数据并行应用中基于拓扑的局部通信网格

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-01 DOI: 10.1109/CLUSTR.2005.347025

S. Figueira, S. Beeby, A. S. Wu

引用次数: 1

Service Augmentation for High End Interactive Data Services 高端交互式数据服务的业务增强

2005 IEEE International Conference on Cluster Computing Pub Date : 2005-09-01 DOI: 10.1109/CLUSTR.2005.347053

M. Wolf, H. Abbasi, Benjamin Collins, David Spain, K. Schwan

{"title":"Service Augmentation for High End Interactive Data Services","authors":"M. Wolf, H. Abbasi, Benjamin Collins, David Spain, K. Schwan","doi":"10.1109/CLUSTR.2005.347053","DOIUrl":"https://doi.org/10.1109/CLUSTR.2005.347053","url":null,"abstract":"Advances in computational science, combined with the increasingly interdisciplinary and geographically distributed research teams, have led to a need to support multi-tiered, data- and meta-data-rich collaboration infrastructures. Our research addresses the interactive, remote tasks undertaken in such collaborations, which require a flexible software infrastructure able to dynamically deploy services where and when needed, and to provide data to clients in the forms in which they require it with suitable levels of end-to-end performance. The concept of service augmentation advanced in this paper seeks to continuously adjust the differences or degrees of incompatibility between the data received and the data displayed or stored by clients. Difference adjustments occur anywhere on the paths between data providers and clients, and compatibility computations leverage all of the resources that may be brought to bear, including CPUs and GPUs on servers and additional data manipulations on server, overlay, and client nodes. A formal structure and experimental evaluations of this concept are performed with the SmartPointer scientific visualization and annotation framework, for which we show that data-driven SLAs provide improved client flexibility and the ability to maintain application-specific notions of quality of service","PeriodicalId":255312,"journal":{"name":"2005 IEEE International Conference on Cluster Computing","volume":"2015 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127754429","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}

引用次数: 23