检索结果-内蒙古大学图书馆

Integrating coordinated checkpointing and recovery mechanisms into DSM synchronization barriers

ACM Journal of Experimental Algorithmics 2006年 11卷 2.9–es页

作者： Boukerche, Azzedine De Melo, Alba Cristina Magalhaes Alves School of Information Technology and Engineering University of Ottawa Canada Department of Computer Science University of Brasilia Brasilia Brazil

distributed shared memory (DSM) creates an abstraction of a physical shared memory that parallel programmers can access. Most recent software DSM systems provide relaxed-memory models that guarantee consistency only at synchronization operations, such as locks and barriers. As the main goal of DSM systems is to provide support for long-term computation-intensive applications, checkpointing and recovery mechanisms are highly desirable. This article presents and evaluates the integration of a coordinated checkpointing mechanism to the barrier primitive that is usually provided with many DSM systems. Our results on some popular benchmarks and a real parallel application show that the overhead introduced during the failure-free execution is often small. © 2006 ACM.

关键词： Algorithms Barrier synchronization D.1.3 [Concurrent Programming]-Parallel Programming D.4.5 [Reliability]-Checkpoint/restart, fault tolerance Design distributed shared memory Performance Reliability

来源：评论

学校读者我要写书评

暂无评论

Selecting threads for workload migration in software distributed shared memory systems

引用

PARALLEL COMPUTING 2002年第6期28卷 893-913页

作者： Liang, TY Shieh, CK Li, JQ Natl Cheng Kung Univ Dept Elect Engn Tainan 70101 Taiwan Leader Univ Dept Comp Sci & Informat Engn Tainan Taiwan

When threads are migrated from heavily loaded nodes to lightly loaded nodes for load balance in software distributed shared memory systems, the communication cost of maintaining data consistency is increased if migration threads are carelessly selected. Program performance is degraded when loss from increased communication exceeds the benefit from load balancing. This study addresses the problem with a novel selection policy called reduction of inter-node sharing costs. The main characteristic of this policy is simultaneously considering thread memory access types and global sharing. The experimental results show that this policy can reduce the communication of benchmark applications by 50% during load balancing. (C) 2002 Elsevier Science B.V. All rights reserved.

关键词： distributed shared memory load balance selection policy memory access types global sharing

来源：评论

学校读者我要写书评

暂无评论

Lazy garbage collection of recovery state for fault-tolerant distributed shared memory (vol 13, pg 673, 2002)

引用

IEEE TRANSACTIONS ON PARALLEL AND distributed SYSTEMS 2002年第10期13卷 1084-1098页

作者： Sultan, F Nguyen, TD Iftode, L Department of Computer Science Rutgers University Piscataway NJ 08854-8019 United States

In this paper, we address the problem of garbage collection in a single-failure fault-tolerant home-based lazy release consistency (HLRC) distributed shared-memory (DSM) system based on independent checkpointing and logging. Our solution uses laziness in garbage collection and exploits consistency constraints of the HLRC memory model for low overhead and scalability. We prove safe bounds on the state that must be retained in the system to guarantee correct recovery after a failure. We devise two algorithms for garbage collection of checkpoints and logs, checkpoint garbage collection (CGC), and lazy log trimming (LLT). The proposed approach targets large-scale distributed shared-memory computing on local-area clusters of computers. In such systems, using global synchronization or extra communication for garbage collection is inefficient or simply impractical due to system scale and temporary disconnections in communication. The challenge lies in controlling the size of the logs and the number of checkpoints without global synchronization while tolerating transient disruptions in communication. Our garbage collection scheme is completely distributed, does not force processes to synchronize, does not add extra messages to the base DSM protocol, and uses only the available DSM protocol information. Evaluation results for real applications show that it effectively bounds the number of past checkpoints to be retained and the size of the logs in stable storage.

关键词： fault tolerance distributed shared memory checkpointing log-based rollback recovery garbage collection

来源：评论

学校读者我要写书评

暂无评论

An optimistic-based partition-processing approach for distributed shared memory systems

引用

JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 2002年第6期18卷 853-869页

作者： Lin, JW Kuo, SY Fu Jen Catholic Univ Dept Comp Sci & Informat Engn Taipei 242 Taiwan Natl Taiwan Univ Dept Elect Engn Taipei 10764 Taiwan

This paper investigates the problem of network partitioning in distributed shared memory (DSM) systems. We propose an optimistic-based partition-processing approach, which can make shared pages available when network partitioning occurs. However, this approach does not guarantee that the same page in different partitions can maintain a consistent value. To eliminate this problem, a memory-based coordinated checkpoininting scheme is presented to save consistent states at low cost. If there are inconsistencies between two partitions, one saved consistent state is chosen to perform backward error recovery. Extensive trace-driven simulations have been performed to evaluate the effects of the proposed approach on system performance.

关键词： network partitioning distributed shared memory checkpointing backward error recovery trace-driven simulations

来源：评论

学校读者我要写书评

暂无评论

Migrating-home protocol for software distributed shared memory

引用

JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 2002年第6期18卷 929-957页

作者： Cheung, BWL Wang, CL Lau, FCM Univ Hong Kong Dept Comp Sci & Informat Syst Hong Kong Hong Kong Peoples R China

The efficiency of Software distributed shared memory (DSM) is often limited by the excessive amount of network communication in maintaining the memory consistency of the system. Two of the most popular software solutions to reduce redundant data traffic are relaxed memory consistency models and traffic-thrifty coherence protocols. In this paper, we propose the migrating-home protocol for a relaxed memory consistency model, the scope consistency model. The protocol allows the processor storing the most up-to-date copy of a page to change from one processor to another, so as to better adapt to the memory access patterns of DSM applications. The new protocol has been implemented in a DSM system running on a 16-node Pentium III 450MHz PC cluster. We analyzed not only the execution time of the benchmark programs, but also the communication and page fault patterns via a new analysis approach. It is shown that our DSM system reduces the amount of network communication and handles page faults more efficiently. The benchmark results provide concrete evidence for the substantial performance improvement obtained by our system.

关键词： clustering computing distributed shared memory memory consistency model coherence protocol home-based protocol migrating-home protocol JUMP

来源：评论

学校读者我要写书评

暂无评论

TRUSS: A reliable, scalable server architecture

引用

IEEE MICRO 2005年第6期25卷 51-59页

作者： Gold, BT Kim, J Smolens, JC Chung, ES Liaskovitis, V Nurvitadhi, E Falsafi, B Hoe, JC Nowatzyk, AG Carnegie Mellon Univ Dept Elect & Comp Engn Pittsburgh PA 15213 USA Cedars Sinai Med Ctr Minimally Invas Surg Technol Inst Los Angeles CA 90048 USA

Traditional techniques that mainframes use to increase reliability special hardware or custom software - are incompatible with commodity server requirements. The truss architecture provides reliable, scalable computat... 详细信息

关键词： WEB servers CLIENT/SERVER computing MULTIPROCESSORS COMPUTER network protocols APPLICATION software MAINFRAME computers COMPUTER architecture SYSTEMS design distributed shared memory

来源：评论

学校读者我要写书评

暂无评论

shared memory computing on clusters with symmetric multiprocessors and system area networks

引用

ACM TRANSACTIONS ON COMPUTER SYSTEMS 2005年第3期23卷 301-335页

作者： Kontothanassis, L Stets, R Hunt, G Rencuzogullari, U Altekar, G Dwarkadas, S Scott, ML Hewlett Packard Corp Cambridge Res Lab HP Labs Cambridge MA 02139 USA Google Inc Mountain View CA 94043 USA Microsoft Corp Res Redmond WA 98052 USA VMware Inc Palo Alto CA 94304 USA Univ Calif Berkeley Dept Elect Engn & Comp Sci Berkeley CA 94720 USA Univ Rochester Dept Comp Sci Rochester NY 14627 USA

Cashmere is a software distributed shared memory (S-DSM) system designed for clusters of server-class machines. It is distinguished from most other S-DSM projects by (1) the effective use of fast user-level messaging, as provided by modern system-area networks, and (2) a "two-level" protocol structure that exploits hardware coherence within multiprocessor nodes. Fast user-level messages change the tradeoffs in coherence protocol design;they allow Cashmere to employ a relatively simple directory-based coherence protocol. Exploiting hardware coherence within SMP nodes improves overall performance when care is taken to avoid interference with inter-node software coherence. We have implemented Cashmere on a Compaq AlphaServer/memory Channel cluster, an architecture that provides fast user-level messages. Experiments indicate that a one-level, version of the Cashmere protocol provides performance comparable to, or slightly better than, that of TreadMarks' lazy release consistency. Comparisons to Compaq's Shasta protocol also suggest that while fast user-level messages make finer-grain software DSMs competitive, VM-based systems continue to outperform software-based access control for applications without extensive fine-grain sharing. Within the family of Cashmere protocols, we find that leveraging intranode hardware coherence provides a 37% performance advantage over a more straightforward one-level implementation. Moreover, contrary to our original expectations, noncoherent hardware support for remote memory writes, total message ordering, and broadcast, provide comparatively little in the way of additional benefits over just fast messaging for our application suite.

关键词： experimentation performance distributed shared memory relaxed consistency software coherence

来源：评论

学校读者我要写书评

暂无评论

Randomized registers and iterative algorithms

引用

distributed COMPUTING 2005年第3期17卷 209-221页

作者： Lee, H Welch, JL Univ Denver Dept Comp Sci Denver CO 80208 USA Texas A&M Univ Dept Comp Sci College Stn TX 77843 USA

We present three different specifications of a read-write register that may occasionally return out-of-date values - namely, a (basic) random register, a P -random register, and a monotone random register. We show that these specifications are implemented by the probabilistic quorum algorithm of Malkhi, Reiter, Wool, and Wright, and we illustrate how to program with such registers in the framework of Bertsekas, using the notation of Uresin and Dubois. Consequently, existing iterative algorithms for a significant class of problems (including solving systems of linear equations, finding shortest paths, constraint satisfaction, and transitive closure) will converge with high probability if executed in a system in which the shared data is implemented with registers satisfying the new specifications. Furthermore, the algorithms in this framework will inherit positive attributes concerning load and fault-tolerance from the underlying register implementation. The expected convergence time for iterative algorithms using the monotone implementation is analyzed and shown experimentally to improve on that of the original implementation. The message complexity for iterative algorithms using the monotone probabilistic quorum implementation is shown to improve on that of non-probabilistic implementations in a quantifiable situation.

关键词： distributed shared memory registers randomization probabilistic quorums iterative algorithms

来源：评论

学校读者我要写书评

暂无评论

In-kernel integration of operating system and infiniband functions for high performance computing clusters: A DSM example

引用

IEEE TRANSACTIONS ON PARALLEL AND distributed SYSTEMS 2005年第9期16卷 830-840页

作者： Liss, L Birk, Y Schuster, A Technion Israel Inst Technol Dept Elect Engn IL-32000 Haifa Israel Technion Israel Inst Technol Dept Comp Sci IL-32000 Haifa Israel

The Infiniband (IB) System Area Network (SAN) enables applications to access hardware directly from user level, reducing the overhead of user-kernel crossings during data transfer. However, distributed applications that exhibit close coupling between network and OS services may benefit from accessing IB from the kernel through IB's native Verbs interface, which permits tight integration of these services. We assess this approach using a sequential-consistency distributed shared memory (DSM) system as an example. We first develop primitives that abstract the low-level communication and kernel details, and efficiently serve the application's communication, memory, and scheduling needs. Next, we combine the primitives to form a kernel DSM protocol. The approach is evaluated using our full-fledged Linux kernel DSM implementation over Infiniband. We show that overheads are reduced substantially, and overall application performance is improved in terms of both absolute execution time and scalability relative to an entirely user level implementation.

关键词： hardware/software interfaces high-speed networks distributed shared memory parallel computing

来源：评论

学校读者我要写书评

暂无评论

Evaluating scheduling policies for fine-grain communication protocols on a cluster of SMPs

引用

JOURNAL OF PARALLEL AND distributed COMPUTING 2005年第4期65卷 464-478页

作者： Falsafi, B Wood, DA Carnegie Mellon Univ Comp Architecture Lab Pittsburgh PA 15213 USA Univ Wisconsin Dept Comp Sci Madison WI 53706 USA

distributed-memory parallel computers and networks of workstations (NOWs) both rely on efficient communication over increasingly high-speed networks. Software communication protocols are often the performance bottleneck. Several current and proposed parallel systems address this problem by dedicating one general-purpose processor in a symmetric multiprocessor (SMP) node specifically for protocol processing. This protocol processing convention reduces communication latency and increases effective bandwidth, but also reduces the peak performance since the dedicated processor no longer performs computation. In this paper, we study a parallel machine with SMP nodes and compare two protocol processing policies: the Fixed policy, which uses a dedicated protocol processor;and the Floating policy, where all processors perform both computation and protocol processing. The results from synthetic microbenchmarks and five macrobenchmarks show that: (i) a dedicated protocol processor benefits light-weight protocols much more than heavy-weight protocols, (ii) a dedicated protocol processor is generally advantageous when there are four or more processors per node, (iii) multiprocessor node performance is not as sensitive to interrupt overhead as uniprocessor node because a message arrival is likely to find an idle processor on a multiprocessor node, thereby eliminating interrupts, (iv) the system with the lowest cost-performance will include a dedicated protocol processor when interrupt overheads are much higher than protocol weight-as in light-weight protocols. (c) 2005 Elsevier Inc. All rights reserved.

关键词： protocol processing policies fine-grain software protocols scheduling communication clusters of SMPs distributed shared memory

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：