检索结果-内蒙古大学图书馆

PetaShare: A reliable, efficient and transparent distributed storage management system

SCIENTIFIC PROGRAMMING 2011年第1期19卷 27-43页

作者： Kosar, Tevfik Akturk, Ismail Balman, Mehmet Wang, Xinqi SUNY Buffalo Dept Comp Sci & Engn Buffalo NY 14260 USA Louisiana State Univ Dept Comp Sci Baton Rouge LA 70803 USA Louisiana State Univ Ctr Computat & Technol Baton Rouge LA 70803 USA Bilkent Univ Dept Comp Engn Ankara Turkey Univ Calif Berkeley Lawrence Berkeley Lab Computat Res Div Berkeley CA 94720 USA

Modern collaborative science has placed increasing burden on data management infrastructure to handle the increasingly large data archives generated. Beside functionality, reliability and availability are also key factors in delivering a data management system that can efficiently and effectively meet the challenges posed and compounded by the unbounded increase in the size of data generated by scientific applications. We have developed a reliable and efficient distributed data storage system, PetaShare, which spans multiple institutions across the state of Louisiana. At the back-end, PetaShare provides a unified name space and efficient data movement across geographically distributed storage sites. At the front-end, it provides light-weight clients the enable easy, transparent and scalable access. In PetaShare, we have designed and implemented an asynchronously replicated multi-master metadata system for enhanced reliability and availability, and an advanced buffering system for improved data transfer performance. In this paper, we present the details of our design and implementation, show performance results, and describe our experience in developing a reliable and efficient distributed data management system for data-intensive science.

关键词： distributed data storage metadata management asynchronous replication advanced buffer reliability performance data-intensive science PetaShare

来源：评论

学校读者我要写书评

暂无评论

Optimized data-Available storage for Energy-Limited Wireless Sensor Networks

Optimized Data-Available Storage for Energy-Limited Wireless...

引用

36th Annual IEEE Conference on Local Computer Networks (LCN)

作者： Kamenik, Jens Theel, Oliver OFFIS Inst Informat Technol Escherweg 2 D-26121 Oldenburg Germany Carl von Ossietzky Univ Oldenburg Dept Comp Sci D-26111 Oldenburg Germany

ISBN: (纸本)9781612849287

Environmental monitoring with wireless sensor networks (WSNs) is based on an infrastructure with unreliable communication channels and nodes. WSNs, also, have a limited lifetime due to the fixed energy budget. Nevertheless, the user of a data-storing WSN often needs high availability of the measured data. But unfortunately, there is no conceptional framework to articulate those QoS requirements nor a design-guideline that leads to a WSN implementation guaranteeing these requirements. In this paper, we address these problems. We present an approach that identifies the organization of the measured data within the WSN giving minimal availability guarantees as demanded by the end-user at minimal energy consumption. We rely on sensor data replication in order to implement query-specific availability guarantees in an energy-optimized manner. We exemplary show the feasibility of our approach through an adaptation of the well-known WSN experiment from Great Duck Island.

关键词： wireless sensor networks distributed data storage data replication energy optimization

来源：评论

学校读者我要写书评

暂无评论

Reliability study of coding schemes for wide-area distributed storage systems

Reliability study of coding schemes for wide-area distribute...

引用

19th International Euromicro Conference on Parallel, distributed and Network-Based Processing (PDP)

作者： Peter, Kathrin Zuse Inst Berlin Comp Sci Res Berlin Germany

ISBN: (纸本)9780769543284

distributed storage systems comprise a large number of commodity hardware distributed across several data centers. Even in the presence of failures (permanent failures) the system should provide reliable storage. While replication has advantages because of its simplicity there exist coding techniques that provide adaptable reliability properties with an optimal redundancy ratio at the same time e.g. MDS (maximum distance separable) erasure codes. The coding and distribution scheme influences the prospective storage reliability. In this paper we present reliability models for erasure coding and replication techniques especially for their application in wide-area storage systems. Furthermore we utilize these models to quantify the reliability properties of concrete data storage scenarios.

关键词： distributed data storage erasure codes reliability Cache Memory System labeling scheme data center Distribute

来源：评论

学校读者我要写书评

暂无评论

tinyDSM: A Highly Reliable Cooperative data storage For Wireless Sensor Networks

tinyDSM: A Highly Reliable Cooperative Data Storage For Wire...

引用

International Symposium on Collaborative Technologies and Systems

作者： Piotrowski, Krzysztof Langendoerfer, Peter Peter, Steffen IHP D-15236 Frankfurt Germany

ISBN: (纸本)9781424445851

The advantage of a Wireless Sensor Network (WSN) compared to a centric approach is the distribution of sensing suites. However in order for such a system of distributed resources to work in a reliable and effective way a smart cooperation between nodes is needed. In this paper we propose a middleware approach for a highly reliable data storage that helps to assure data availability despite the well known WSN resource problems and disappearing or inactive nodes by providing a reasonable data redundancy in the system. Such a solution helps to ease the design and optimization of the data exchange between nodes as well. Our solution is configurable in order to satisfy the needs of the application on top regarding performance/requirements trade-off. The options specify the quantity and quality of the data replication. Additional features like event mechanism that monitors the data and the possibility to issue database like queries increase the applicability of our middleware. In this paper we focus on the evaluation of its capabilities regarding reliability, the consistency of replicates and the costs of the data management. The simulation results for a reasonable set-up show that the CPU load caused by the data replication is low (below 3 percent) and the average inconsistency time is as small as about 0,06 seconds for a single hop and about 0, 15 seconds for a two hops replication area. There is still room for improvements, but a clear definition of problems helps to find ways to cope with them in order to achieve the chosen goals.

关键词： Cooperative Computing Wireless Sensor Networks Software Development distributed Shared Memory distributed data storage

来源：评论

学校读者我要写书评

暂无评论

DISP: Practical, Efficient, Secure and Fault-Tolerant distributed data storage

引用

ACM Transactions on storage 2005年第1期1卷 71-94页

作者： Ellard, Daniel Megquier, James Sun Microsystems MA 1 Network Drive Burlington 01803 United States etc consulting Japan

DISP is a practical client-server protocol for the distributed storage of immutable data objects. Unlike most other contemporary protocols, DISP permits applications to make explicit tradeoffs between total storage space, computational overhead, and guarantees of availability, integrity, and privacy on a per-object basis. Applications specify the degree of redundancy with which each item is encoded, what level of integrity checks are computed and stored with each item, and whether items are stored in an encrypted format. At one extreme, clients willing to pay the overhead are guaranteed privacy, integrity, and availability of data stored in the system as long as fewer than half the servers are Byzantine. At the other extreme, objects that do not require privacy or integrity in the face of Byzantine servers can be stored with very low computational and storage overhead. DISP is efficient in terms of message count, message size, and storage requirements: even in the worst case, the read and write protocols require a number of messages that are linear with respect to the number of servers. In terms of message size, DISP requires transferring only marginally more than L bytes to correctly read an object of size L, even in the face of Byzantine server failures. In this article we provide a description of DISP and an analysis of its fault-tolerant properties. We also analyze the complexity of the protocol and discuss several potential applications. We conclude with a description of our prototype implementation and measurements of its performance on commodity hardware. © 2005, ACM. All rights reserved.

关键词： Design distributed data storage Reliability

来源：评论

学校读者我要写书评

暂无评论

A peer-to-peer infrastructure for distributing large scientific data sets over wide area high-performance networks: experimental studies using wide area layer 2 services 07

A peer-to-peer infrastructure for distributing large scienti...

引用

Proceedings of the first international conference on Networks for grid applications

作者： Yunhong Gu Robert L. Grossman Joe Mambretti University of Illinois at Chicago Northwestern University

ISBN: (纸本)9789639799028

This paper presents Sector, a distributed environment that was created specifically to address the challenges inherent in accessing, exploring, analyzing and transporting extremely large scientific datasets over high performance wide area networks. To date, the effective utilization of such datasets has been limited because accessing and transporting large remote data sets in traditional distributed computing environments is often a challenge. Sector was designed and developed to eliminate these barriers. The Sector architecture incorporates specialized communications services and specialized data services that are designed for high volume data flows over wide area high performance optical networks. This design employs several innovative techniques to ensure that data flows are maximized at all times and at all locations required regardless of distances. This architectural design has been implemented on a prototype international experimental testbed. The results and implications of these preliminary tests are described.

关键词： high performance networks distributed data storage

来源：评论

学校读者我要写书评

暂无评论

An approach for robust distributed data retrieval in data Intensive Grid environments

An approach for robust distributed data retrieval in Data In...

引用

1st International Symposium on Pervasive Computing and Applications

作者： Ye, Wei Gu, Ning Fudan Univ Dept Comp & Informat Technol Shanghai 200433 Peoples R China

ISBN: (纸本)1424403251

data Intensive Grid is meant for efficient data sharing in large scale distributed data storage environment, which is a promising way to improve data and software viability. Among numerous aspects which needs consideration, data availability is one of the most vital and difficult, since it faces lots of challenges such as abnormal failures of nodes, dynamic nodes participation and unstable network connection. Adding data replicas in the data Grid environment can improve data availability and response time and thus ensure robust data retrieval. Yet how to replicate data in order to meet performance requirements in data Grid still remains unsolved. In this paper we propose a solution that stripes large files into data blocks and distributes over data nodes for storage and design a mechanism for automatic data replication of nodes aiming at ensuring data availability. Preliminary simulation experiments show that our approach is effective in improving data availability with limited increase of system costs.

关键词： distributed data storage grid replication data retrieval

来源：评论

学校读者我要写书评

暂无评论

The Roma personal metadata service

引用

MOBILE NETWORKS & APPLICATIONS 2002年第5期7卷 407-418页

作者： Swierk, E Kscsman, E Williams, NC Fukushima, T Yoshida, H Laviano, V Baker, M Stanford Univ Dept Comp Sci Stanford CA 94305 USA

People now have available to them a diversity of digital storage facilities, including laptops, cell phone address books, handheld devices, desktop computers and web-based storage services. Unfortunately, as the number of personal data repositories increases, so does the management problem of ensuring that the most up-to-date version of any document in a user's personal file space is available to him on the storage facility he is currently using. We introduce the Roma personal metadata service to make it easier to locate current versions of personal files and ensure their availability across different repositories. This centralized service stores information about each of a user's files, such as name, location, timestamp and keywords, on behalf of mobility-aware applications. Separating out these metadata from the data respositories makes it practical to keep the metadata store on a highly available, portable device. In this paper we describe the design requirements, architecture and current prototype implementation of Roma.

关键词： metadata mobile computing distributed databases distributed data storage personal systems data synchronization

来源：评论

学校读者我要写书评

暂无评论

Compact, adaptive placement schemes for non-uniform requirements 02

Compact, adaptive placement schemes for non-uniform requirem...

引用

Proceedings of the fourteenth annual ACM symposium on Parallel algorithms and architectures

作者： André Brinkmann Kay Salzwedel Christian Scheideler University of Paderborn Paderborn Germany Johns Hopkins University Baltimore MD

ISBN: (纸本)9781581135299

In this paper we study the problem of designing compact, adaptive strategies for the distribution of objects among a heterogeneous set of servers. Ideally, such a strategy should allow the computation of the position of an object with a low time and space complexity, and it should be able to adapt with a near-minimum amount of replacements of objects to changes in the capabilities of the servers so that objects are always distributed among the servers according to their capabilities. Previous techniques are able to handle these requirements only in part. For example, standard hashing techniques can be used to achieve a non-uniform distribution of objects among a set of servers and the time and space efficient computation of the position of the objects, but they usually do not adapt well to a change in the capabilities. We present two strategies based on hashing that achieve all of the goals above. Furthermore, we give a list of applications for these strategies demonstrating that they can be used efficiently for distributed data management, web caches, and adaptive random graphs, which may be of interest for peer-to-peer networks.

关键词： adaptive hashing random graphs balls into bins RAID web caching non-uniform disks distributed data storage

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：