Distributed Processing Chapter 1 : Introduction

1
Distributed ProcessingChapter 1 Introduction
2
Problem

• There are n nodes, each of which has a value. A
  node wants to know the maximum value among the n
  nodes.
• Centralized Approach
• A server maintains the values of n nodes and each
  node reports its value to the server.
• Then the query node sends a message to ask the
  maximum value to the server, which will answer to
  the query.
• Distributed Approach
• Each node communicates with its 6 nearest
  neighbor nodes to inform its value.
• Then the query node eventually finds the maximum
  value by exchanging information with its neighbor
  nodes.

3
Discussion

• Question 1 Find the algorithm for distributed
  approach.
• Question 2 Compare the performance
• In terms of the number of communications
• Question 3 Make a comparison table for the two
  approaches

4
Definition of a Distributed System

• Distributed system
• 1) A collection of (scalability)
• 2) independent computers that (heterogeneity)
• 3) appears to its users as a single coherent
  system (transparency)
• Distributed System versus Parallel System
• Separated Operating System vs. Single Operating
  System
• Message Passing vs. Shared Memory

5
Why Distributed System ?

• Performance
• Incremental Growth (Scalability)
• 1 single mainframe of price W
• N small machines of price W/N
• Fault Tolerance
• 1 single mainframe critical weak point
• Failure of a machine replacement by other
  machines
• Geographical Distribution and Availability
• Flexible configuration
• e.g. 1 Disk server, 3 Computing servers, 1
  Graphic server, etc.
• Geographical availability

6
Distributed System - Scalibility and
Heterogeneity
1.1
A distributed system organized as middleware. ?
Heterogeneity and Scalability
7
Distributed System - Transparency
Different forms of transparency in a distributed
system.
Transparency Description
Access Hide differences in data representation and how a resource is accessed
Location Hide where a resource is located
Migration Hide that a resource may move to another location
Relocation Hide that a resource may be moved to another location while in use
Replication Hide that a resource may be shared by several competitive users
Concurrency Hide that a resource may be shared by several competitive users
Failure Hide the failure and recovery of a resource
Persistence Hide whether a (software) resource is in memory or on disk
8
Distributed System Heterogeneity
Application Program or Client
Client has to be provided with one different
driver for each server
9
Distributed System Heterogeneity and
Object-Oriented Approach
Application Program or Client
Predefined interface
Wrapping with predefined interface
Encapsulation Object-Oriented Approach
10
Hardware Concepts Multiprocessor
1.6
11
Multiprocessors (1)

• A bus-based multiprocessor.

1.7
12
Multiprocessors (2)

• (a) A crossbar switch
• (b) An omega switching network

1.8
13
Homogeneous Multicomputer Systems

• (a) Grid
• (b) Hypercube

14
Software Concepts
System Description Main Goal
DOS Tightly-coupled operating system for multi-processors and homogeneous multicomputers Hide and manage hardware resources
NOS Loosely-coupled operating system for heterogeneous multicomputers (LAN and WAN) Offer local services to remote clients
Middleware Additional layer atop of NOS implementing general-purpose services Provide distribution transparency

• An overview of
• DOS (Distributed Operating Systems)
• NOS (Network Operating Systems)
• Middleware

15
Issues in System Design

• Transparency
• Flexibility
• Reliability
• Performance
• Scalability
• Interoperability

16
Transparency

• Hiding physical details about
• Location
• Migration
• Duplication
• Relocation
• Concurrency
• Parallelism
• Location
• Access

17
Flexibility

• Should be easy to modify functionality and
  architecture
• To provide with Configurability, Avalability and
  Autonomy
• Micro-Kernel vs. Monolithic Kernel
• Monolithic Kernel Provides all functionalities
  of OS. example. UNIX
• Micro-Kernel
• Minimal subset of OS what users want
• Example
• Kernel Watch

18
Reliability

• Important Goal of Distributed System
• Reliability
• Security
• Fault-Tolerance
• Failure Probability P
• Should be P P1P2P3 Pn
• But often P P1 P2 P3 Pn in reality

19
Performance and Scalability

• Improve performance by parallelism
• Throughput T
• Ideally should be T Tn when n is the number of
  sites
• In reality T lt Tn
• Due to some Bottleneck

Throughput
Number of sites
??
20
Granularity of Parallelism

• Unit of Task
• Fine-Granularity vs. Coarse Granularity
• Fine-Granularity
• Large number of small tasks
• Need a large amount of inter-task communication
• Not good for distributed system (good for
  Parallel system)
• Coarse-Granularity
• Small number of big tasks
• Only small amount of inter-task communication
• Good for distributed system

21
Interoperability

• Easy to collaborate with other systems in
  run-time
• Compatibility, Portability
• How to achieve Interoperability
• Well-Defined API set
• Standardization

22
Hardware Concepts Multiprocessor
1.6
23
Multiprocessors (1)

• A bus-based multiprocessor.

1.7
24
Multiprocessors (2)

• (a) A crossbar switch
• (b) An omega switching network

1.8
25
Homogeneous Multicomputer Systems

• (a) Grid
• (b) Hypercube