Dynamicity in distributed applications: issues, problems and the ASSIST approach

Dynamic reconfiguration refers to the ability of changing properties and structure of a distributed system, while it is running. It is essential when designing applications that need to adapt to unpredictable events, run on non-dedicated environments; for which configuration cannot be statically determined, or may change at runtime. Grid computing also gave new emphasis to the topic, being based on dynamic environments by definition. The first part of this work tries to give a definite organization to known literature and existing experiments, providing a general overview of the problem. Then we describe our approach to dynamicity in the context of the ASSIST programming environment and show how it is able to provide users with many of the required abstractions to develop adaptive, high-performance, distributed applications.     

A Decentralized Computational Infrastructure for Grid-Based Parallel Asynchronous Iterative Applications

Parallel asynchronous iterative algorithms relax synchronization and communication requirements, and can potentially extend Desktop Grids beyond embarrassingly parallel applications to support a broader class of parallel iterative applications. This paper presents the design and implementation of CometG, a decentralized (peer-to-peer) computational infrastructure that extends Desktop Grid environments to support these applications. CometG provides a decentralized and scalable tuple space, efficient communication and coordination support, and application-level abstractions that can be used to implement Desktop Grid applications based on parallel asynchronous iterative algorithms using the master-worker/BOT paradigm. The deployment and evaluations of CometG and a CometG-based application in a wide-area environment using the PlanetLab [7] test bed, as well as a campus network are presented.     

考虑网络性能的广域网系统中机群的通信模型

网格中的机群或者超级计算机通过广域网互相连接，在这个平台上进行并行编程应用的一个主要问题是它们的等级网络结构，广域网上的延迟和带宽通常是局域网中的好几倍。该文针对LogP模型进行了扩展，提出带参数的LogP模型，详细讨论比较了其中的各个参数以及如何通过实验对它们进行测量。在此模型基础上通过选取合适的通信结构，并且把消息分割成多个部分。在不同的广域连接上采用并行化发送，达到优化通信操作的目的。     

A slowdown model for applications executing on time-shared clustersof workstations

Distributed applications executing on clustered environments typically share resources (computers and network links) with other applications. In such systems, application execution may be retarded by the competition for these shared resources. In this paper, we define a model that calculates the slowdown imposed on applications in time-shared multi-user clusters. Our model focuses on three kinds of slowdown: local slowdown, which synthesizes the effect of contention for CPU in a single workstation; communication slowdown, which synthesizes the effect of contention for the workstations and network links on communication costs; and aggregate slowdown, which determines the effect of contention on a parallel task caused by other applications executing on the entire cluster, i.e., on the nodes used by the parallel application. We verify empirically that this model provides an accurate estimate of application performance for a set of compute-intensive parallel applications on different clusters with a variety of emulated loads     

Automatic Flow-Control Adaptation for Enhancing Network Performance in Computational Grids

With the advent of computational Grids, networking performance over the wide-area network (WAN) has become a critical component in the Grid infrastructure. Unfortunately, many high-performance Grid applications only use a small fraction of their available bandwidth because operating systems and their associated protocol stacks are still tuned for yesterday's WAN speeds. As a result, network gurus undertake the tedious process of manually tuning system buffers to allow TCP flow control to scale to today's WAN Grid environments. And although recent research has shown how to set the size of these system buffers automatically at connection set-up, the buffer sizes are only appropriate at the beginning of the connection's lifetime. To address these problems, we describe an automated and lightweight technique called dynamic right-sizing that can improve throughput by as much as an order of magnitude while still abiding by TCP semantics.     

A case study of Web server benchmarking using parallel WAN emulation

This paper describes the use of a parallel discrete-event network emulator called the Internet Protocol Traffic and Network Emulator (IP-TNE) for Web server benchmarking. The experiments in this paper demonstrate the feasibility of high-performance wide area network (WAN) emulation using parallel discrete-event simulation (PDES) techniques on a single shared-memory multiprocessor. Our experiments with an Apache Web server achieve up to 8000 HTTP/1.1 transactions/s for static document retrieval across emulated WAN topologies with up to 4096 concurrent Web/TCP clients. The results show that WAN characteristics, including round-trip delays, packet losses, and bandwidth asymmetry, all have significant impacts on Web server performance, as do client protocol behaviors. WAN emulation using the IP-TNE enables stress testing and benchmarking of Web servers in ways that may not be possible in simple local area network (LAN) test scenarios.     

A Hybrid Analysis of an Optimization Approach for Cluster Applications

Cluster/distributed computing has become a popular, cost-effective alternative to high-performance parallel computers. Many parallel programming languages and related programming models have become widely accepted on clusters. However, the high communication overhead is a major shortcoming of running parallel applications on cluster/distributed computing environments. To reduce the communication overhead and thus the completion time of a parallel application, this paper introduces and evaluates an efficient Key Message (KM) approach to support parallel computing on cluster computing environments. In this paper, we briefly present the model and algorithm, and then analytical and simulation methods are adopted to evaluate the performance of the algorithm. It demonstrates that when network background load increases or the computation to communication ratio decreases, the analysis results show better improvement on communication of a parallel application over the system which does not use the KM approach.     

DP: a paradigm for anonymous remote, computation and communicationfor cluster computing

This paper explores the transparent programmability of communicating parallel tasks in a Network of, Workstations (NOW). Programs which are tied up with specific machines will not be resilient to the changing conditions of a NOW. The Distributed Pipes (DP) model enables location independent intertask communication among processes' across machines. This approach enables migration of communicating parallel tasks according to runtime conditions. A transparent programming model for a parallel solution to Iterative Grid Computations using DP is also proposed. Programs written using the model are resilient to the heterogeneity of nodes and changing conditions in the NOW. They are also devoid of any network related code. The design of runtime support and function library support are presented. An engineering problem, namely, the Steady State Equilibrium Problem, is studied over the model. The performance analysis shows the speedup due to parallel execution and scaled down memory requirements. We present a case where the effect of communication overhead can be nullified to achieve a linear to super-linear speedup. The analysis discusses performance resilience of Iterative Grid Computations and characterizes synchronization delay among subtasks and the effect of network overhead and load fluctuations on performance. The performance saturation characteristics of such applications are also studied     

Ninf-G: A Reference Implementation of RPC-based Programming Middleware for Grid Computing

GridRPC, which is an RPC mechanism tailored for the Grid, is an attractive programming model for Grid computing. This paper reports on the design and implementation of a GridRPC programming system called Ninf-G. Ninf-G is a reference implementation of the GridRPC API which has been proposed for standardization at the Global Grid Forum. In this paper, we describe the design, implementations and typical usage of Ninf-G. A preliminary performance evaluation in both WAN and LAN environments is also reported. Implemented on top of the Globus Toolkit, Ninf-G provides a simple and easy programming interface based on standard Grid protocols and the API for Grid Computing. The overhead of remote procedure calls in Ninf-G is acceptable in both WAN and LAN environments.     

Performance of parallel computations with dynamic processor allocation

In parallel adaptive mesh refinement (AMR) computations the problem size can vary significantly during a simulation. The goal here is to explore the performance implications of dynamically varying the number of processors proportional to the problem size during simulation. An emulator has been developed to assess the effects of this approach on parallel communication, parallel runtime and resource consumption. The computation and communication models used in the emulator are described in detail. Results using the emulator with different AMR strategies are described for a test case. Results show for the test case, varying the number of processors, on average, reduces the total parallel communications overhead from 16 to 19% and improves parallel runtime time from 4 to 8%. These results also show that on average resource utilization improves more than 37%.     

基于Windows平台的数据包实时传输模拟

采用基于Windows平台的网络驱动程序接口规范技术,设计并实现一个基于协议过滤的高精度网络模拟器。该模拟器可实时模拟广域网传输链路的各种网络传输条件,包括网络带宽、数据包的丢包、传输延迟及传输错序等。性能测试结果表明,该网络模拟器达到了高精度、低负荷的设计要求。其实现有助于研究多媒体网络应用服务质量、进行网络协议分析以及验证各种网络流量控制算法。     

Distributed Downloads of Bulk,Replicated Grid Data

Data-sharing scientific communities use storage systems as distributed data stores by replicating content. In such highly replicated environments, a particular dataset can reside at multiple locations and can thus be downloaded from any one of them. Since datasets of interest are significantly large in size, improving download speeds either by server selection or by co-allocation can offer substantial benefits. In this paper, we present an architecture for co-allocating Grid data transfers across multiple connections, enabling the parallel download of datasets from multiple servers. We have developed several co-allocation strategies comprising of simple brute-force, predictive and dynamic load balancing techniques as a means both to exploit rate differences among the various client–server links and to address dynamic rate fluctuations. We evaluate our approaches using the GridFTP data movement protocol in a wide-area testbed and present our results.     

基于对象的可靠群组通信机制研究

近年来,网格计算技术日益成为用来解决数据和计算密集型应用的可行方案,网格运行平台本身和在网格环境中的并行应用都需要大量的点对多点的群组通信.提出一种灵活、可容错的群组通信机制.该机制基于远程方法调用(RMI),可为分布式并行应用提供高效、可容错的群组通信.通信方法可以在本地对象、远程对象,或一组对象中激活.这种通信采用异步方式,通信发起者可以选择全等待或必要性等待两种机制来获取通信结果.从而最大程度地保证通信的可靠性或高效性.     

AMROEBA: Computational astrophysics modeling enabled by dynamic lambda switching

Many data and compute intensive Grid applications, such as computational astrophysics, may be able to benefit from networking supported by dynamically provisioned lightpaths. To date, the majority of high performance distributed environments have been based on traditional routed packet networks, provisioned as external services rather than as integrated components within those environments. Because this approach often cannot provide high performance capabilities required by these applications, an alternative distributed infrastructure architecture is being designed based on dynamic lightpaths, supported by optical networks. These designs implement communication services and infrastructure as integral components of distributed infrastructure. The resultant environments resemble large scale specialized instruments. Presented here is one such architecture, implemented on a wide-area, optical Grid test bed, featuring a closely integrated dedicated lightpath mesh. The test bed was used to conduct a series of experiments to explore its potential for supporting adaptive mesh refinement (AMR) astrophysics simulations. While preliminary, the results of these experiments indicate that this architecture may provide the deterministic capabilities required by a wide range of high performance distributed services and applications, especially for computational science.     

Dynamic Instrumentation, Performance Monitoring and Analysis of Grid Scientific Workflows

While existing work concentrates on developing QoS models of business workflows and Web services, few tools have been developed to support the monitoring and performance analysis of scientific workflows in Grids. This paper describes novel Grid services for dynamic instrumentation of Grid-based applications, performance monitoring and analysis of Grid scientific workflows. We describe a Grid dynamic instrumentation service that provides a widely accessible interface for other services and users to conduct the dynamic instrumentation of Grid applications during the runtime. We introduce a Grid performance analysis service for Grid scientific workflows. The analysis service utilizes various types of data including workflow graphs, monitoring data of resources, execution status of activities, and performance measurements obtained from the dynamic instrumentation of invoked applications, and provides a rich set of functionalities and features to support the online monitoring and performance analysis of scientific workflows. Workflows and their relevant information including performance metrics are stored and utilized for comparing the performance of constructs of different workflows and for supporting multi-workflow analysis. The work described in this paper is supported in part by the Austrian Science Fund as part of the Aurora Project under contract SFBF1104 and by the European Union through the IST-2002-511385 project K-WfGrid.     

Grid programming: some indications where we are headed

Grid computing enables the development of large scientific applications on an unprecedented scale. Grid-aware applications, also called meta-applications or multi-disciplinary applications, make use of coupled computational resources that are not available at a single site. In this light, the Grids let scientists solve larger or new problems by pooling together resources that could not be coupled easily before. It is well known that the programmer’s productivity in designing and implementing efficient distributed/parallel applications on high-performance computers is still usually a very time-consuming task. Grid computing makes the situation worse. Consequently, the development of Grid programming environments that would enable programmers to efficiently exploit this technology is an important and hot research issue.

After an introduction on the main Grid programming issues, this paper will review the most important approaches/projects conducted in this field worldwide.     

Local search to improve coordinate-based task mapping

We present a local search strategy to improve the coordinate-based mapping of a parallel job’s tasks to the MPI ranks of its parallel allocation in order to reduce network congestion and the job’s communication time. The goal is to reduce the number of network hops between communicating pairs of ranks. Our target is applications with a nearest-neighbor stencil communication pattern running on mesh systems with non-contiguous processor allocation, such as Cray XE and XK Systems. Using the miniGhost mini-app, which models the shock physics application CTH, we demonstrate that our strategy reduces application running time while also reducing the runtime variability. We further show that mapping quality can vary based on the selected allocation algorithm, even between allocation algorithms of similar apparent quality.     

一种基于CCM构件模型的计算网格平台

如何聚合网络中分布异构的计算资源来解决大规模的科学计算问题,和如何减少并行程序设计的复杂性,一直是网格计算研究的难点之一。文章提出了一种基于CORBA构件技术的计算网格新思想,构造了一个计算网格的模型(CCGM)。该模型能够充分地利用构件技术带来的可组装和易管理的特性来形成网格计算。并通过问题的抽象定义和使用ParIDL工具将问题的定义映射到CCGM之上,简化了计算网格应用的开发。通过测试和分析CCG(ComponentbasedComputationalGrid)系统,表明CCG系统具有较好的加速比。     

一个网格应用程序监测系统的实现

并行程序在运行过程中会受到外部运行环境的干扰而出现错误,并行程序调试可以帮助用户发现并排除程序中存在的错误。只有对程序的运行进行监测才能获得调试数据,并对监测到的数据进行分析,以达到调试程序的目的。网格计算的提出推动了并行计算的发展,越来越多的应用程序在网格环境中运行。该文从以上两点出发,开发了一个网格应用程序的监测系统,可以用以调试网格上的应用程序。