检索结果-内蒙古大学图书馆

Integrating relational databases with support for updates 88

proceedings of the first international symposium on databases in parallel and distributed systems

作者： M. Samy Gamal-Eldin G. Thomas R. Elmasri MCS Department Clarkson University Potsdam NY Bellcore 444 Hoes Lane Piscataway NJ Computer Science Department University of Houston at University Park Houston TX

ISBN: (纸本)9780818608933

Most work on database integration has considered only support for data retrieval, not support for updates, and often the use of a special semantically rich data model has been required. In this paper we present an approach to database integration which supports updates and which uses only the standard relational data model. Many of the ideas used in this approach are applicable to database integration in the context of other data models as well.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Effect of skew on join performance in parallel architectures 88

引用

proceedings of the first international symposium on databases in parallel and distributed systems

作者： M. Seetha Lakshmi P. S. Yu IBM Research Division T.J. Watson Research Center P.O. Box 704 Yorktown Heights N.Y.

ISBN: (纸本)9780818608933

Skew in the distribution of values taken by an attribute is identified as a major factor that can affect the performance of parallel architectures for relational joins. The effect of skew on the performance of two parallel architectures is evaluated using analytic models. In one architecture, called database machine (DBMC), data as well as processing power are distributed; while in the other architecture, called Single Processor parallel Input/output (SPPI), data is distributed but the processing power is concentrated in one processor. The two architectures are compared in terms of the ratio of MIPS used by DBMC and SPPI to deliver the same throughput and response time. In addition, the horizontal growth potential of DBMC is evaluated in terms of maximum speedup achievable by DBMC relative to SPPI response time. The MIPS ratio as well as speedup are found to be very sensitive to the amount of skew. These suggest, careful thought should be given in parallelizing database applications and in the design of algorithms and query optimizer for parallel architectures.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Job scheduling that minimizes network contention due to both communication and I/O

Job scheduling that minimizes network contention due to both...

引用

international symposium on parallel and distributed Processing (IPDPS)

作者： J. Mache V. Lo S. Garg Lewis-Clark State College University of Oregon Intel Corporation

As communication and I/O traffic increase on the interconnection network of high-performance systems, network contention becomes a critical problem drastically reducing performance. Whereas earlier allocation strategies were either sensitive to communication alone or sensitive to I/O alone, we present a new strategy that is sensitive to both communication and I/O. Our new strategy MC-Elongated, strives to achieve (1) the compactness needed to minimize communication-based contention as well as (2) the balance and orientation relative to I/O nodes needed to minimize I/O-based contention. We tested our new strategy using synthetic workloads and a real workload trace of 6087 jobs captured from a 400 node Intel Paragon. Our results show that with respect to system throughput and average job turnaround time, in environments with varying degree of communication and I/O traffic, MC-Elongated outperforms previous allocation strategies that are in use today. Regarding the tension between communication and I/O, our results show that spatial layout is more critical for I/O intensive jobs at lower utilization levels and more critical for communication-intensive jobs at higher utilization levels; and that in general, the impact of I/O traffic is dominant.

关键词： Telecommunication traffic Testing Multimedia databases Educational institutions Ear Read only memory Data visualization Visual databases Multiprocessor interconnection networks Degradation

来源：评论

学校读者我要写书评

暂无评论

An efficient algorithm for processing distributed queries using partition dependency

An efficient algorithm for processing distributed queries us...

引用

international Conference on parallel and distributed systems (ICPADS)

作者： H. Chen C. Liu School of Computer Science Telecommunications and Information Systems De Paul University Chicago IL USA

Presents an efficient algorithm for processing distributed queries with the existence of partition dependencies. For a given query, the algorithm first partitions the referenced relations into a number of non-exclusive subsets such that the join operation(s) associated with the relations in the subset can be locally processed without data transfer. Each subset is associated with a set of processing sites and can be used to generate an execution plan for the given query. Then, the algorithm determines a set of referenced fragmented relations that are not in the subset, such that only the fragments (instead of the whole relation) need to be replicated at the processing sites. The other referenced relations are duplicated at each of the processing sites. Among the alternatives, the algorithm picks the plan that gives the minimum response time for the query. Experimental results show that our algorithm improves the performance of distributed query processing significantly.

关键词： Partitioning algorithms Query processing Qualifications Computer science Information systems Delay databases Computer architecture Disk drives Random access memory

来源：评论

学校读者我要写书评

暂无评论

Prophesy: an infrastructure for analyzing and modeling the performance of parallel and distributed applications

Prophesy: an infrastructure for analyzing and modeling the p...

引用

international symposium on High Performance distributed Computing

作者： V.E. Taylor X. Wu J. Geisler X. Li Z. Lan R. Stevens M. Hereld I.R. Judson Department of Electrical and Computer Engineering Celero Research Group Northwestern University Evanston IL USA MCS Division Argonne National Laboratory Argonne IL USA

Efficient execution of applications requires insight into how the system features impact the performance of the application. For distributed systems, the task of gaining this insight is complicated by the complexity of the system features. This insight generally results from significant experimental analysis and possibly the development of performance models. This paper presents the Prophesy project, an infrastructure that aids in gaining this needed insight based upon experience. The core component of Prophesy is a relational database that allows for the recording of performance data, system features and application details.

关键词： Performance analysis Data analysis Application software Analytical models Instruments Relational databases Kernel Scalability Algorithm design and analysis Laboratories

来源：评论

学校读者我要写书评

暂无评论

Optimal periodic remapping of bulk synchronous computations on multiprogrammed distributed systems

Proceedings of the International Parallel Processing Symposi...

引用

proceedings of the international parallel Processing symposium, IPPS 2000年 103-108页

作者： Fong, Ngo-Tai Xu, Cheng-Zhong Wang, Le Yi Wayne State Univ Detroit United States

For bulk synchronous computations that have non-deterministic behaviors, dynamic remapping is an effective approach to ensure parallel efficiency. There are two basic issues in remapping: when and how to remap. This paper presents a formal treatment of the first issue for dynamic computations with a priori known statistical behaviors. We have formulated the problem as two complement sequential stochastic optimization, with an objective of finding optimal remapping frequencies for a given tolerance of load imbalance on multiprogrammed distributed systems. We have developed analytical approaches to precisely characterize the transient statistical behaviors of the workload process and derived optimal remapping frequencies for various random workload change processes.

关键词： parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

Support for recoverable memory in the distributed virtual communication machine

Proceedings of the International Parallel Processing Symposi...

引用

proceedings of the international parallel Processing symposium, IPPS 2000年 191-198页

作者： Rosu, Marcel-Catalin Schwan, Karsten IBM T.J. Watson Research Cent Yorktown Heights United States

The distributed Virtual Communication Machine (DVCM) is a software communication architecture for clusters of workstations equipped with programmable network interfaces (NIs) for high-speed networks. DVCM is an extensible architecture, which promotes the transfer of application modules to the NI. By executing `closer' to the network, on the NI CoProcessor, these modules can communicate with significantly higher message rates and lower latencies than achievable at the CPU-level. This paper describes how DVCM modules can be used to enhance the performance of the Cluster Recoverable Memory system (CRMem), a transaction-processing kernel for memory-resident databases. By using the NI CoProcessor for CRMem's remote operations, our implementation achieves more than 3,000 trans/sec on a simplified TpcB benchmark.

关键词： distributed computer systems

来源：评论

学校读者我要写书评

暂无评论

A robust protocol for parallel join operation in distributed data bases 88

引用

proceedings of the first international symposium on databases in parallel and distributed systems

作者： S. Bandyopadhyay A. Sengupta School of Computer Science University of Windsor Windsor Ont N9B 3P4 CANADA Dept of Computer Science University of South Carolina Columbia SC USA

ISBN: (纸本)9780818608933

Fault tolerant distributed databases use replicated data(e.g., record or relation) to handle failures of one or more nodes in a computer network. Efficient and economic access strategies for such data bases have not been investigated. In this paper, the binary hypercube, a popular model for fault tolerant interconnection networks, has been studied. It has been shown that, for a local area network based on a binary hypercube, having 2r nodes where every data is replicated r times, in the absence of faults, any query involving an arbitrary sequence of joins R1 @@@@ R2 @@@@ … @@@@ Rn, n ≤ r, may be performed by repeatedly executing joins in a distributed fashion using n node disjoint paths from n/2 distinct sites of database operations to n arbitrary sites containing a target relation each. This protocol also solves the problem of materialization of relations. In the presence of up to r-2 faults, the protocol still guarantees N node disjoint paths to arbitrary sites. The value of N is determined by the number of faulty nodes.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Using postordering and static symbolic factorization for parallel sparse LU

Using postordering and static symbolic factorization for par...

引用

international symposium on parallel and distributed Processing (IPDPS)

作者： M. Cosnard L. Grigori LORIA INRIA-Lorraine Nancy France LORIA-Universite Henri Poincare-Nancy I Nancy France

In this paper we present several improvements of widely used parallel LU factorization methods on sparse matrices. first we introduce the LU elimination forest and then we characterize the L, U factors in terms of their corresponding LU elimination forest. This characterization can be used as a compact storage scheme of the matrix as well as of the task dependence graph. To improve the use of BLAS in the numerical factorization, we perform a postorder traversal of the LU elimination forest, thus obtaining larger supernodes. To expose more task parallelism for a sparse matrix, we build a more accurate task dependence graph that includes only the least necessary dependences. Experiments compared favorably our methods against methods implemented in the S* environment on the SGI's Origin2000 multiprocessor.

关键词： Sparse matrices Microwave integrated circuits Ice Linear systems Concurrent computing parallel processing Packaging machines Scalability Differential equations Tree data structures

来源：评论

学校读者我要写书评

暂无评论

Self-stabilizing mutual exclusion using unfair distributed scheduler

Self-stabilizing mutual exclusion using unfair distributed s...

引用

international symposium on parallel and distributed Processing (IPDPS)

作者： A.K. Datta M. Gradinariu S. Tixeuil Department of Computer Science University of Nevada-Las Vegas Laboratoire de Recherche en Informatique Universite de Paris Sud France

A self-stabilizing algorithm, regardless of the initial system state, converges infinite time to a set of states that satisfy a legitimacy predicate without the need for explicit exception handler of backward recovery. Mutual exclusion is fundamental in the area of distributed computing, by serializing the accesses to a common shared resource. All existing probabilistic self-stabilizing mutual exclusion algorithms designed to work under an unfair distributed scheduler suffer from the following common drawback: Once stabilized, there exists no upper bound of time between two executions of the critical section at a given node. We present the first probabilistic self-stabilizing algorithm that guarantees such a bound (O(n/sup 3/), where n is the network size) while working using an unfair distributed scheduler. As the scheduling adversary gets weaker the bound gets better. Our algorithm works in an anonymous unidirectional ring of any size and has a O(n/sup 3/) expected stabilization time.

关键词： Read only memory Fault tolerant systems Processor scheduling Computer science Electronic switching systems Military computing Upper bound Electrical capacitance tomography Humans

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：