Distributed Computing Software based solutions to Parallel Computing.

Slides:



Advertisements
Similar presentations
Distributed Processing, Client/Server and Clusters
Advertisements

Distributed Systems CS
Chorus and other Microkernels Presented by: Jonathan Tanner and Brian Doyle Articles By: Jon Udell Peter D. Varhol Dick Pountain.
Distributed Systems 1 Topics  What is a Distributed System?  Why Distributed Systems?  Examples of Distributed Systems  Distributed System Requirements.
PERFORMANCE TESTING OF PVM AND MPI WITH THE MOSIX PROCESS MIGRATION IN A DISTRIBUTED COMPUTING ENVIRONMENT Mr. Ye Myint Naing
CS-550: Distributed File Systems [SiS]1 Resource Management in Distributed Systems: Distributed File Systems.
Using DSVM to Implement a Distributed File System Ramon Lawrence Dept. of Computer Science
Objektorienteret Middleware Presentation 2: Distributed Systems – A brush up, and relations to Middleware, Heterogeneity & Transparency.
Distributed Processing, Client/Server, and Clusters
Software Tools for Dynamic Resource Management Irina V. Shoshmina, Dmitry Yu. Malashonok, Sergay Yu. Romanov Institute of High-Performance Computing and.
Transparent Process Migration for Distributed Applications in a Beowulf Cluster Mark Claypool and David Finkel Computer Science Department Worcester Polytechnic.
Tanenbaum & Van Steen, Distributed Systems: Principles and Paradigms, 2e, (c) 2007 Prentice-Hall, Inc. All rights reserved DISTRIBUTED SYSTEMS.
Threads Clients Servers Code Migration Software Agents Summary
Scripting Languages For Virtual Worlds. Outline Necessary Features Classes, Prototypes, and Mixins Static vs. Dynamic Typing Concurrency Versioning Distribution.
1 Introduction to Load Balancing: l Definition of Distributed systems. Collection of independent loosely coupled computing resources. l Load Balancing.
16: Distributed Systems1 DISTRIBUTED SYSTEM STRUCTURES NETWORK OPERATING SYSTEMS The users are aware of the physical structure of the network. Each site.
1 Distributed Systems: Distributed Process Management – Process Migration.
Hardware/Software Concepts Tran, Van Hoai Department of Systems & Networking Faculty of Computer Science & Engineering HCMC University of Technology.
DISTRIBUTED PROCESS IMPLEMENTAION BHAVIN KANSARA.
Grid Monitoring By Zoran Obradovic CSE-510 October 2007.
Design and Implementation of a Single System Image Operating System for High Performance Computing on Clusters Christine MORIN PARIS project-team, IRISA/INRIA.
Distributed Process Implementation Hima Mandava. OUTLINE Logical Model Of Local And Remote Processes Application scenarios Remote Service Remote Execution.
 Introduction Introduction  Definition of Operating System Definition of Operating System  Abstract View of OperatingSystem Abstract View of OperatingSystem.
Distributed File Systems Concepts & Overview. Goals and Criteria Goal: present to a user a coherent, efficient, and manageable system for long-term data.
Ch4: Distributed Systems Architectures. Typically, system with several interconnected computers that do not share clock or memory. Motivation: tie together.
Computer Architecture Parallel Processing
Reference: / Parallel Programming Paradigm Yeni Herdiyeni Dept of Computer Science, IPB.
SSI-OSCAR A Single System Image for OSCAR Clusters Geoffroy Vallée INRIA – PARIS project team COSET-1 June 26th, 2004.
Chapter 9 Message Passing Copyright © Operating Systems, by Dhananjay Dhamdhere Copyright © Operating Systems, by Dhananjay Dhamdhere2 Introduction.
UNIX System Administration OS Kernal Copyright 2002, Dr. Ken Hoganson All rights reserved. OS Kernel Concept Kernel or MicroKernel Concept: An OS architecture-design.
Chapter 6 Operating System Support. This chapter describes how middleware is supported by the operating system facilities at the nodes of a distributed.
Submitted by: Shailendra Kumar Sharma 06EYTCS049.
Socket Swapping for efficient distributed communication between migrating processes MS Final Defense Praveen Ramanan 12 th Dec 2002.
Transparency in Distributed Operating Systems Vijay Akkineni.
Loosely Coupled Parallelism: Clusters. Context We have studied older archictures for loosely coupled parallelism, such as mesh’s, hypercubes etc, which.
Processes and Threads Processes have two characteristics: – Resource ownership - process includes a virtual address space to hold the process image – Scheduling/execution.
Tanenbaum & Van Steen, Distributed Systems: Principles and Paradigms, 2e, (c) 2007 Prentice-Hall, Inc. All rights reserved DISTRIBUTED SYSTEMS.
Copyright © George Coulouris, Jean Dollimore, Tim Kindberg This material is made available for private study and for direct.
Windows NT Operating System. Windows NT Models Layered Model Client/Server Model Object Model Symmetric Multiprocessing.
Types of Operating Systems
Comparison of Distributed Operating Systems. Systems Discussed ◦Plan 9 ◦AgentOS ◦Clouds ◦E1 ◦MOSIX.
DISTRIBUTED SYSTEMS Principles and Paradigms Second Edition ANDREW S
PART II OPERATING SYSTEMS LECTURE 8 SO TAXONOMY Ştefan Stăncescu 1.
Issues Autonomic operation (fault tolerance) Minimize interference to applications Hardware support for new operating systems Resource management (global.
Derek Wright Computer Sciences Department University of Wisconsin-Madison MPI Scheduling in Condor: An.
Processes Introduction to Operating Systems: Module 3.
The Cosmic Cube Charles L. Seitz Presented By: Jason D. Robey 2 APR 03.
CORBA1 Distributed Software Systems Any software system can be physically distributed By distributed coupling we get the following:  Improved performance.
Distributed database system
Tanenbaum & Van Steen, Distributed Systems: Principles and Paradigms, 2e, (c) 2007 Prentice-Hall, Inc. All rights reserved DISTRIBUTED SYSTEMS.
Definition of a Distributed System (1) A distributed system is: A collection of independent computers that appears to its users as a single coherent system.
Types of Operating Systems 1 Computer Engineering Department Distributed Systems Course Assoc. Prof. Dr. Ahmet Sayar Kocaeli University - Fall 2015.
Department of Computer Science and Software Engineering
Parallel IO for Cluster Computing Tran, Van Hoai.
FTOP: A library for fault tolerance in a cluster R. Badrinath Rakesh Gupta Nisheeth Shrivastava.
Background Computer System Architectures Computer System Software.
SYSTEM MODELS FOR ADVANCED COMPUTING Jhashuva. U 1 Asst. Prof CSE
Chapter 16 Client/Server Computing Dave Bremer Otago Polytechnic, N.Z. ©2008, Prentice Hall Operating Systems: Internals and Design Principles, 6/E William.
INTRODUCTION TO HIGH PERFORMANCE COMPUTING AND TERMINOLOGY.
PVM and MPI.
Enhancements for Voltaire’s InfiniBand simulator
DISTRIBUTED SYSTEMS Principles and Paradigms Second Edition ANDREW S
File System Implementation
Introduction to Load Balancing:
OpenMosix, Open SSI, and LinuxPMI
Parallel Virtual Machine
University of Technology
Auburn University COMP7500 Advanced Operating Systems I/O-Aware Load Balancing Techniques (2) Dr. Xiao Qin Auburn University.
Distributed System Structures 16: Distributed Structures
Outline Midterm results summary Distributed file systems – continued
Presentation transcript:

Distributed Computing Software based solutions to Parallel Computing

Distributed Computing b Cluster based computing b Network of Workstations Duality of standalone workstation Vs. parallel computing environment.Duality of standalone workstation Vs. parallel computing environment. b Distributed Applications RC5 ClientsRC5 Clients SETI ClientsSETI Clients

Issues of Distributed Computing b Homogeneous networks Networks of similar machines. Requires only one version of executables.Networks of similar machines. Requires only one version of executables. b Heterogeneous networks Different Architectures of machines. (Alpha,x86]. Each architecture requires its own version of executables.Different Architectures of machines. (Alpha,x86]. Each architecture requires its own version of executables. b Issues of data representation, system level interface differences, and suitability.

Support Model b Application support is critcal. b Three basic models Application is built with full awareness and supportApplication is built with full awareness and support User land facilities are provided to facilitate awareness and supportUser land facilities are provided to facilitate awareness and support Kernel level supportKernel level support

Application Support b Application modeled as self-contained entity capable of client-server, peer-peer cooperation models. RC5 clientRC5 client SETI clientSETI client b Support can be streamlined b Loss of transparency, load balancing,lower QoS.

User Level Support b Provide basic parallel mechanisms Message/Data passingMessage/Data passing SynchronizationSynchronization Process HandlingProcess Handling –Rough control b Provides greater transparency Less dependence on architecture, transparency supports heterogeneous solutionsLess dependence on architecture, transparency supports heterogeneous solutions

Kernel Support b Provides system level support SchedulingScheduling –Load Balancing –Process Migration Resource AllocationResource Allocation b Doesn’t work well in heterogeneous configurations. b Process Migration increases communication costs

Three Solutions b PVM - Parallel Virtual Machine User Land Parallel SolutionUser Land Parallel Solution Message PassingMessage Passing b MPI - Message Passing Interface StandardStandard Several ImplementationsSeveral Implementations User Land SupportUser Land Support

Three Solutions - continued b MOSIX - Linux Kernel Extensions Provide dynamic schedulingProvide dynamic scheduling process migrationprocess migration system level integrationsystem level integration

PVM - Parallel Virtual Machine b Message Passing Paradigm b Heterogeneous Computing Deals effectively with data representation issuesDeals effectively with data representation issues b Supports variety of systems including MPP’s, SMP’s, and vector machines b Uses a daemon to provide parallel facilities

PVM - continued b Tracks tasks using a unique system assigned ID called a TID b Supports grouping, and group level activities.

MPI or LAM b Message Passing Interface b MPI is the standard, several implementations exist. One is LAM and is maintained by Notre Dame. b Similar to PVM HeterogeneousHeterogeneous Uses a daemon, also includes a peer-peer mode.Uses a daemon, also includes a peer-peer mode.

MPI - LAM - continued b LAM environment must be started explicitly b Provides compiler shells to handle program compilation b Nodes are dynamic b Suite of utilities to maintain the message passing virtual machine. b The parallelism is explicitly programmed

MPI - LAM - continued b MPI addresses resource limitations. Uses a property called Guaranteed Envelope Resources to maintain integrity of processes.

MOSIX and Linux b Linux, not much more to say. b Mosix is developed to extend several unix operating systems b Homogeneous b Transparent and Preemptive process migration b Dynamic process reassignment

Mosix - continued b Global resource assignment b Ideal for cluster based computing b Offers memory ushering prevents thrashingprevents thrashing b Maintains home system b Requires that the nodes of the cluster be networked well

Mosix - Continued b Mosix maintains a given process’ association with a home node. Unique home nodeUnique home node b Implements a bi-level approach to maintain this association through migration User Context (remote) - system independent contextsUser Context (remote) - system independent contexts

Mosix - continued Deputy - node dependent interface. Remains when process migrates from home node.Deputy - node dependent interface. Remains when process migrates from home node. b All system dependent calls are routed through remote to deputy at home system gettimeofdaygettimeofday b Initial assignment needs to be done by PVM or MPI. Mosix doesn’t handle this aspect.

MOSIX - Migrated Process

Bad Mosix b Overhead is increased Delayed system callsDelayed system calls I/O sockets (file, network) don’t route wellI/O sockets (file, network) don’t route well Migration time adds overheadMigration time adds overhead b Needs further work in migratable sockets and files

You’re a good audience b Are you asleep yet. b But I tried! b Thank you, b Any Questions?