检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Wei, Xingda Lu, Fangming Chen, Rong Chen, Haibo Institute of Parallel and Distributed Systems SEIEE Shanghai Jiao Tong University China Shanghai AI Laboratory China

We present KRCORE, an RDMA library with a microsecond-scale control plane on commodity RDMA hardware for elastic computing. KRCORE can establish a full-fledged RDMA connection within 10µs (hundreds or thousands of times faster than verbs), while only maintaining a (small) fixed-sized connection metadata at each node, regardless of the cluster scale. The key ideas include virtualizing pre-initialized kernel-space RDMA connections instead of creating one from scratch, and retrofitting advanced RDMA dynamic connected transport with static transport for both low connection overhead and high networking speed. Under load spikes, KRCORE can shorten the worker bootstrap time of an existing disaggregated key-value store (namely RACE Hashing) by 83%. In serverless computing (namely Fn), KRCORE can also reduce the latency for transferring data through RDMA by 99%. © 2021, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Numerical analysis of small-strain elasto-plastic deformation using local Radial Basis Function approximation with Picard iteration

arXiv

引用

arXiv 2024年

作者： Strniša, Filip Jančič, Mitja Kosec, Gregor "Jožef Stefan" Institute Parallel and Distributed Systems Laboratory Jamova cesta 39 Ljubljana1000 Slovenia

This paper deals with a numerical analysis of plastic deformation under various conditions, utilizing Radial Basis Function (RBF) approximation. The focus is on the elasto-plastic von Mises problem under plane-strain assumption. Elastic deformation is modelled using the Navier-Cauchy equation. In regions where the von Mises stress surpasses the yield stress, corrections are applied locally through a return mapping algorithm. The non-linear deformation problem in the plastic domain is solved using the Picard iteration. The solutions for the Navier-Cauchy equation are computed using the Radial Basis Function-Generated Finite Differences (RBF-FD) meshless method using only scattered nodes in a strong form. Verification of the method is performed through the analysis of an internally pressurized thick-walled cylinder subjected to varying loading conditions. These conditions induce states of elastic expansion, perfectly-plastic yielding, and plastic yielding with linear hardening. The results are benchmarked against analytical solutions and traditional Finite Element Method (FEM) solutions. The paper also showcases the robustness of this approach by solving case of thick-walled cylinder with cut-outs. The results affirm that the RBF-FD method produces results comparable to those obtained through FEM, while offering substantial benefits in managing complex geometries without the necessity for conventional meshing, along with other benefits of meshless methods. © 2024, CC BY.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

The parallelized Cuckoo Filter for Cold Data Representation 23

The Parallelized Cuckoo Filter for Cold Data Representation

引用

23rd IEEE International Conference on High Performance Computing and Communications, 7th IEEE International Conference on Data Science and systems, 19th IEEE International Conference on Smart City and 7th IEEE International Conference on Dependability in Sensor, Cloud and Big Data systems and Applications, HPCC-DSS-SmartCity-DependSys 2021

作者： Sun, Bowen Luo, Lailong Li, Shangsen Chen, Yingwen Guo, Deke College of Computer National University of Defense Technology China National University of Defense Technology Science and Technology on Information Systems Engineering Laboratory China College of Computer National University of Defense Technology National Lab for Parallel and Distributed Processing China

ISBN: (纸本)9781665494571

Cold data contributes a large portion of the big data today and is usually stored in secondary storage. Various sketch data structures are implemented to represent the stored elements and provide constant-time membership queries. Considering the large volume and infrequent access of cold data, we envision three rationales of cold data representation: 1) parallelized insertion, 2) accurate representation, and 3) constant-time query/deletion. The existing sketches, however, fail to achieve these rationales simultaneously. To this end, we propose a novel sketch named parallelized cuckoo Filter (PCF) in this paper. PCF divides the standard cuckoo filter into multiple homogenous segments. An additional hash function is introduced to assign an element to a segment. The element will be inserted into the selected segment. With such a design, element insertions in PCF can be naturally parallelized. Besides, as a variant of the cuckoo filter, PCF naturally realizes the accurate representation and constant-time query/deletion requirements. Comprehensive experiments further indicate that PCF achieves the three rationales simultaneously. © 2021 IEEE.

关键词： Hash functions Smart cities Big Data Data structures Standards

来源：评论

学校读者我要写书评

暂无评论

Characterizing Off-path SmartNIC for Accelerating distributed systems

arXiv

引用

arXiv 2022年

作者： Wei, Xingda Cheng, Rongxin Yang, Yuhan Chen, Rong Chen, Haibo Institute of Parallel and Distributed Systems SEIEE Shanghai Jiao Tong University China Shanghai AI Laboratory China

SmartNICs have recently emerged as an appealing device for accelerating distributed systems. However, there has not been a comprehensive characterization of SmartNICs, and existing designs typically only leverage a single communication path for workload offloading. This paper presents the first holistic study of a representative off-path SmartNIC, specifically the Bluefield-2, from a communication-path perspective. Our experimental study systematically explores the key performance characteristics of communication among the client, on-board SoC, and host, and offers insightful findings and advice for designers. Moreover, we propose the concurrent use of multiple communication paths of a SmartNIC and present a pioneering guideline to expose new optimization opportunities for various distributed systems. To demonstrate the effectiveness of our approach, we conducted case studies on a SmartNIC-based distributed file system (LineFS) and an RDMA-based disaggregated key-value store (DrTM-KV). Our experimental results show improvements of up to 30% and 25% for LineFS and DrTM-KV, respectively. © 2022, CC BY.

关键词： System-on-chip

来源：评论

学校读者我要写书评

暂无评论

No Provisioned Concurrency: Fast RDMA-codesigned Remote Fork for Serverless Computing

arXiv

引用

arXiv 2022年

作者： Wei, Xingda Lu, Fangming Wang, Tianxia Gu, Jinyu Yang, Yuhan Chen, Rong Chen, Haibo Institute of Parallel and Distributed Systems SEIEE Shanghai Jiao Tong University China Shanghai AI Laboratory China

Serverless platforms essentially face a tradeoff between container startup time and provisioned concurrency (i.e., cached instances), which is further exaggerated by the frequent need for remote container initialization. This paper presents MITOSIS, an operating system primitive that provides fast remote fork, which exploits a deep codesign of the OS kernel with RDMA. By leveraging the fast remote read capability of RDMA and partial state transfer across serverless containers, MITOSIS bridges the performance gap between local and remote container initialization. MITOSIS is the first to fork over 10,000 new containers from one instance across multiple machines within a second, while allowing the new containers to efficiently transfer the pre-materialized states of the forked one. We have implemented MITOSIS on Linux and integrated it with FN, a popular serverless platform. Under load spikes in real-world serverless workloads, MITOSIS reduces the function tail latency by 89% with orders of magnitude lower memory usage. For serverless workflow that requires state transfer, MITOSIS improves its execution time by 86%. © 2022, CC BY.

关键词： Containers

来源：评论

学校读者我要写书评

暂无评论

A risk-free multiversion temporally correct algorithm

A risk-free multiversion temporally correct algorithm

引用

International Symposium on parallel and distributed Processing (IPDPS)

作者： A. Boukerche T. Tuck Department of Computer Sciences Parallel Simulation and Distributed Systems (PARADISE) Research Laboratory University of North Texas USA

来源：评论

学校读者我要写书评

暂无评论

A distributed synchronization scheme for multimedia streams in mobile systems: proof and correctness

A distributed synchronization scheme for multimedia streams ...

引用

Conference on Local Computer Networks (LCN)

作者： A. Boukerche Sungbum Hong T. Jacob Parallel Simulation and Distributed Systems (PARADISE) Research Laboratory Department of Computer Science University of North Texas USA

ISBN: (纸本)0769513212

The rapid development in portable computing platforms, distributed multimedia systems and wireless communication technologies has led to a significant interest in mobile multimedia and wireless networking. Video entertainment and multimedia Internet access over wireless networks are examples of future wireless personal communication systems. These, however, incur challenging technological issues. The synchronization of the multimedia objects among mobile clients is one of the most important of these issues. The task of synchronization in wireless multimedia systems is to eliminate variations and delays incurred during the transmission of multiple-media streams and to maintain the temporal order among the various media streams. We propose an efficient distributed multimedia synchronization algorithm for wireless and mobile multimedia systems to ensure and facilitate mobile clients' access to multimedia objects. We discuss the algorithm and provide its formal proof and correctness.

关键词： Streaming media Multimedia systems Wireless communication Computer networks distributed computing Portable computers Communications technology Mobile computing Multimedia computing IP networks

来源：评论

学校读者我要写书评

暂无评论

An algorithm for guaranteeing synchronous ordering of messages

An algorithm for guaranteeing synchronous ordering of messag...

引用

International Symposium on Autonomous Decentralized systems (ISADAS)

作者： V.V. Murty V.K. Garg Parallel and Distributed Systems Laboratory Department of Electrical & Computer Engineering University of Texas Austin Austin TX USA

The paper studies the characteristics of synchronous ordering of messages. Synchronous ordering of messages defines synchronous communication based on the causality rather than time. We present the sufficient conditions, based on the causality relations, for any algorithm to provide synchronous ordering. We also propose an algorithm using acknowledgment messages to implement the sufficient conditions. The algorithm is deadlock-free, and provides a higher degree of concurrency than existing algorithms.< >

关键词： Sufficient conditions System recovery Concurrent computing

来源：评论

学校读者我要写书评

暂无评论

Deriving distributed algorithms from a general predicate detector

Deriving distributed algorithms from a general predicate det...

引用

IEEE Annual International Computer Software and Applications Conference (COMPSAC)

作者： J.R. Mitchell V.K. Garg Parallel and Distributed Systems Laboratory Electrical and Computer Engineering Department University of Texas Austin Austin TX USA

Designing and debugging distributed systems requires the detection of conditions across the entire system. As an illustration, monitoring the status of an application requires detection of termination, and using virtual time requires the periodic calculation of the global virtual time. The generalized conjunctive predicate (GCP) detector offers a method to derive detection algorithms for these and other problems based on optimizing the base algorithm.

关键词： distributed algorithms Detectors distributed computing Debugging Concurrent computing Design engineering Computerized monitoring Application software Design optimization Algorithm design and analysis

来源：评论

学校读者我要写书评

暂无评论

A simulation based study of on-demand routing protocols for ad hoc wireless networks

A simulation based study of on-demand routing protocols for ...

引用

Annual Symposium on Simulation

作者： A. Boukerche Parallel Simulation and Distributed Systems PARADISE Research Laboratory Department of Computer Science University of North Texas USA

ISBN: (纸本)0769510922

Ad hoc networks are wireless, mobile networks that can be set up anywhere and anytime without the aid of any established infrastructure or centralized administration. Because of the limited range of each host's wireless transmission, to communicate with hosts outside its transmission range, a host needs to enlist the aid of its nearby hosts in forwarding packets to the destination. However, since there is no stationary infrastructure such as base stations, each host has to act as a router for itself. A routing protocol for ad hoc networks is executed on every host and is therefore subject to the limit of the resources at each mobile host. A good routing protocol should minimize the computing load on the host as well as the traffic overhead on the network. Therefore, a number of routing protocols have been proposed for ad hoc wireless networks. We focus upon on-demand schemes. We study and compare the performance of the following three routing protocols AODV, CBRP and DSR. A variety of workload and scenarios, as characterized by mobility, load and size of the ad hoc network were simulated. Our results indicate that despite its improvement in reducing route request packets, CBRP has a higher overhead than DSR because of its periodic hello messages while AODV's end-to-end packet delay is the shortest when compared to DSR and CBRP.

关键词： Routing protocols Wireless networks Ad hoc networks Computational modeling Computer simulation Wireless communication Base stations Computer networks Laboratories Computer science

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：