Introduction to Distributed Systems

Chapter 1

1
[email protected]
Course Outline
1. Introduction to DS: Introduction and Definition, Characteristics, Goals of a
Distributed System, Types of Distributed Systems.
2. Architectures: Architectural Styles, System Architectures.
3. Processes: Threads and their Implementation, Anatomy of Clients, Servers and
Design Issues, Code Migration
4. Communication: Network Protocols and Standards, Remote Procedure Call,
Message-Oriented Communication, Stream-Oriented Communication, Multicast
Communication.
5. Naming: Names, Identifiers, and Addresses, Flat Naming, Structured Naming,
Attribute-Based Naming,
6. Synchronization: Clock Synchronization, Logical Clocks, Mutual Exclusion,
Election algorithms,
7. Consistency and Replication: Reasons for Replication, Data-Centric
Consistency Models, Client-Centric Consistency Models, Replica Management,
Consistency Protocols,
8. Fault Tolerance: Introduction to Fault Tolerance, Process Resilience, Reliable
Client-Server Communication, Reliable Group Communication, Distributed
Commit, Recovery
2
Course Outline Continued
 The main Objective of this course is to introduce:
 Explain what a distributed system is
 The current DS development Techniques
 Their construction issues
 Issues that are involved in building reliable
distributed systems, and
 Possible applications of distributed systems.
 Design a distributed system that fulfills
requirements with regards to key distributed
systems properties.
3
 Introduction
 Before the mid-80s:computers were
 Very expensive (hundred of thousands or even millions of
dollars).
 Very slow (a few thousand instructions per second).
 Not connected among themselves.
 After the mid-80s: two major developments
 Cheap and powerful microprocessor-based computers
appeared.
 Computer networks
 LANs at speeds ranging from 10 to 1000 Mbps (now even
10Gbps)
 WANs at speed ranging from 64 Kbps to gigabits/sec.

 Consequence
 Feasibility of using a large network of computers to work for
4the same application; this is in contrast to the old centralized
systems where there was a single computer with its
 Definition of a Distributed System
 A distributed system is a collection of independent
computers that appears to its users as a single
coherent system computer (Tanenbaum & Van Steen)
 This definition has two aspects:
 hardware: autonomous machines
 software: a single system view for the users
 A distributed system is one that stops you getting
any work done when a machine you have never even
heard of crashes (Leslie)
 A distributed system is one in which components
located at networked computers communicate and
coordinate their actions only by passing messages.
(George C.)
5
Middleware
 The middleware layer extends over multiple
machines, and offers each application the
same interface.
 Goal is to hide the heterogeneity of the
underlying OS and HWs
 What does it contain?
 Commonly used components and functions
that need not be implemented by
applications separately.

6
Middleware
 Middleware is software that usually referred as the OS of
distributed systems
 Manage resources for its applications.
 Moreover, it offers services that can also be found in most operating
systems, including:
 Facilities for inter-application communication.
 Security services.
 Accounting services.
 Masking of and recovery from failures.
 Example middleware service.
 Remote Procedure Call (RPC) - allows an
application to invoke a function that is
implemented and executed on a remote computer
as if it was locally available.
7
Examples of Distributed Systems
 Local Area Network
 Database Management System
 Automatic Teller Machine Network
 Internet/World-Wide Web
 Mobile Computing/Mobile
Communication

8
Local Area Network
email server Desktop
computers
print and other servers

Local area
Web server network

email server
print
File server
other servers
the rest of
the Internet
router/firewall

9
Database Management System

10
Automatic Teller Machine

11
Mobile Computing/Mobile
Communication

12
 Characteristics of Distributed Systems
 Differences between the computers and the ways
they communicate are hidden from users
 Users and applications can interact with a
distributed system in a consistent and uniform
way regardless of location
 Distributed systems should be easy to expand
and scale
 A distributed system is normally continuously
available, even if there may be partial failures
(Fault tolerance)

13
 Advantages of Distributed Systems
 Performance: Very often a collection of processors can
provide higher performance (and better
price/performance ratio) than a centralized computer.
 Distribution: many applications involve, by their nature,
spatially separated machines (banking, commercial,
automotive system).
 Reliability (fault tolerance): if some of the machines
crash, the system can survive.
 Incremental growth: as requirements on processing
power grow, new machines can be added incrementally.
 Sharing of data/resources: shared data is essential to
many applications (banking, computer supported
cooperative work, reservation systems); other resources
can be also shared (e.g. expensive printers).
 Communication:
14 facilitates human-to-human
 Disadvantages of Distributed Systems
 Difficulties of developing distributed software:
how should operating systems, programming
languages and applications look like?
 Networking problems: several problems are
created by the network infrastructure, which have to
be dealt with: loss of messages, overloading, ...
 Security problems: sharing generates the problem
of data security.

15
Goals of a Distributed System -Making Resources
accessible
 Connect users and resources.
 Easily connect (printers, computers, storage
facilities, data, files, Web pages, ...)..Some of the
reasons..
 economics: sharing resources such as printers and
high-speed computers..
 to collaborate and exchange information
 groupware: software for collaborative editing,
teleconferencing, etc.
 e-commerce: buying and selling goods.

16
 Goals of Distributed System - Transparency
 A distributed system that is able to present itself
to users and applications as if it were only a
single computer system is said to be
transparent.

17
Different forms of transparency in a distributed
system

18
 Goals of Distributed System - Openness
 Openness is concerned with extensions and
improvements of distributed systems according to
standard rules that describe their syntax and
semantics.
 Detailed interfaces of components need to be
published.
 New components have to be integrated with existing
components.
 Differences in data representation of interface types on
different processors (of different vendors) have to be
resolved.
 Interoperability: components of different origin can
communicate
 Portability: components work on different platforms
19
 Goals of Distributed System - Scalability
 A distributed system should be scalable; there are
three dimensions
 size: adding more users and resources to the system
 geographically: users and resources may be far
apart
 administratively: should be easy to manage even if
it spans many administrative organizations
 Scalability allows the system and applications to
expand in scale without change to the system
structure or the application algorithms.
 But a scalable system
Scalability may exhibit performance
Problems
Concept
problems Example
Centralized services - A single server for all users
Centralized data - A single on-line telephone book
Centralized algorithms - Doing routing based on
complete
20 information
 Scaling Techniques: How to solve scaling
problems
 The problem is mainly performance, and arises as a
result of limitations in the capacity of servers and
networks (for geographical scalability with high
latency and mostly unreliable links)
 Three possible solutions:
 Hiding communication latencies
 Distribution and
 Replication

21
A. Hide Communication Latencies

22
 B. Distribution
 Means splitting a component into smaller parts and
spreading those parts across the system
 e.g., DNS -Domain Name System ..divide the name space
into non over lapping zones

23
 C. Replication
 Replicate components across a distributed system to
increase availability and for load balancing, leading
to better performance
 replication is decided by the owner of a resource
 Caching (a special form of replication) also reduces
communication latency; decided by the user
 but, caching and replication may lead to consistency
problems

24
 Pitfalls when Developing Distributed Systems
 Because of false assumptions made by first time
developers (of distributed systems) which are related
to the properties of distributed systems and do not
occur in non distributed applications.
 The network is reliable (making it difficult to
achieve failure transparency)
 The network is secure
 The network is homogeneous
 The topology does not change
 Latency is zero
 Bandwidth is infinite
 Transport cost is zero
 There is one administrator
25
 Types of Distributed System
 Three main types:
 Distributed computing systems
 Focus on computation
 Goal: High performance computing tasks
 Distributed information systems
 Focus on interoperability (the ability to exchange and
use information)
 Goal: Distribute information across several servers
 Distributed pervasive systems
 Focus on mobile, embedded, communicating systems
 Goal: Spread a real-life environment with a large variety
of smart devices

26
Distributed Computing Systems: Cluster Computing
 Essentially a group of systems connected through a high
speed LAN.
 Homogeneous: Same OS, near-identical hardware
 Single managing node
 Tightly coupled systems
 A master node runs a middleware (containing libraries
for parallel programs) and controls other compute nodes
 Centralized job management & scheduling system

27
 Distributed Computing Systems: Grid Computing
 Lots of nodes (including clusters across multiple subnets) from
everywhere.
 Heterogeneous: no assumptions are made concerning
hardware, operating systems, networks, administrative
domains, security policies, etc.
 Diversity and dynamism (it can handle nodes dropping in
and out at any point of time)
 Dispersed across several organizations and can easily span
a wide-area network
 To allow for collaborations, grids generally use virtual
organizations (grouping of users that will allow for
authorization on resource allocation).
 Loosely coupled (decentralization)
 Distributed job management & scheduling

28
Distributed Computing Systems: Cloud Computing
 Web-based tools or applications that users can access and
use through a web browser as if it were a program
installed locally on their own computer.
 Internet-based computing
 offers dynamically scalable and virtualized resources
that make up services for users to use over the internet
 The only thing the user's computer needs to be able to
run is the cloud computing system's interface software

29
 Distributed Information Systems
 Evolved in organizations that were confronted with a wealth of
networked applications, but for which interoperability turned out
to be problematic.
 Many of the existing middleware solutions are the result of
working with an infrastructure in which it was easier to integrate
applications into an enterprise-wide information system. Several
levels at which integration took place:
 A networked application simply consisted of a server running that
application (often including a database) and making it available to
remote programs, called clients.
 Such clients could send a request to the server for executing a
specific operation, after which a response would be sent back.
Integration at the lowest level would allow clients to wrap a
number of requests, possibly for different servers, into a single
larger request and have it executed as a distributed transaction.
 The key idea was that all, or none of the requests would be
executed.
30
 … Cont’d
 As applications became more sophisticated and were gradually separated into
independent components, it became clear that integration should also take place
by letting applications communicate directly with each other.
 The vast amount of distributed systems in use today is in the form of traditional
information systems.
 Example: Transaction processing systems
BEGIN TRANSACTION(server, transaction);
READ(transaction, file-1, data);
WRITE(transaction, file-2, data);
newData := MODIFIED(data);
IF WRONG(newData) THEN
ABORT TRANSACTION(transaction);
ELSE
WRITE(transaction, file-2, newData);
END TRANSACTION(transaction);
END IF;
Note:
 All READ and WRITE operations are executed, i.e. their effects are made
permanent at the execution of END TRANSACTION.
 Transactions form an atomic operation.
31
Distributed Information Systems: Transaction processing systems
 Focus on database applications - operations on a database
are usually carried out in the form of transactions.
 special primitives are required to program transactions,
supplied either by the underlying distributed system or by the
language runtime system
 Exact list of primitives depends on the type of application;
procedure calls, ordinary statements, etc. can also be
included

32
 … Cont’d
 Transaction between processes:
 e.g., assume the following banking operation
 withdraw an amount x from account 1
 deposit the amount x to account 2
 what happens if there is a problem after the first activity
is carried out?
 group the two operations into one transaction; either
both are carried out or neither
 we need a way to roll back when a transaction is not
completed

33
… Cont’d
 A transaction is a collection of operations on the state of an
object (database, object composition, etc.) that satisfies the
following properties (ACID):
 Atomicity: All operations either succeed, or all of them fail.
 When the transaction fails, the state of the object will
remain unaffected by the transaction.
 Consistency: A transaction establishes a valid state
transition.
 This does not exclude the possibility of invalid,
intermediate states during the transaction’s execution.
 Isolation: Concurrent transactions do not interfere with each
other.
 It appears to each transaction T that other transactions
occur either before T, or after T, but never both.
 Durability: After the execution of a transaction, its effects
34
are made permanent:
 Distributed Pervasive Systems
 The distributed systems discussed so far are characterized
by their stability; fixed nodes having high-quality
connection to a network
 A next-generation of distributed systems emerging in
which the nodes are small, wireless, battery-powered,
mobile (e.g. PDAs, smart phones, wireless surveillance
cameras, portable ECG monitors, etc.), and often
embedded as part of a larger system.
 Some requirements:
 Contextual change: The system is part of an
environment in which changes should be immediately
accounted for.
 Ad hoc composition: Each node may be used in a very
different ways by different users.
 Requires ease-of-configuration.
35
 Sharing is the default: Nodes come and go, providing
 Distributed Pervasive Systems: Examples
 Home Systems
 built around home networks·
 consist of one or more personal computers
 integrate consumer electronics such as TVs, audio and video
equipment, gaming devices, (smart) phones, PDAs, and other
personal wearables into a single system.
 Now/Soon: all kinds of devices such as kitchen appliances,
surveillance cameras, clocks, controllers for lighting, and so
on, will all be hooked up into a single distributed system.
 Several challenges:
 System should be completely self-configuring and self-
managing
 e.g. Universal Plug and Play (UPnP) standards by which
devices automatically obtain IP addresses, can discover each
other, etc.
 Unclear how software and firmware in devices can be easily
updated without manual intervention, or when updates do take
36
 Distributed Pervasive Systems: Examples
Electronic health System
 New devices are being developed to monitor the well-being of
individuals and to automatically contact physicians when
needed.
 Major goal is to prevent people from being hospitalized.
 Questions to be raised:
 Where and how should monitored data be stored?
 How can we prevent loss of crucial data?
 What infrastructure is needed to generate and propagate alerts?
 How can security be enforced?
 How can physicians provide online feedback?

37
 Distributed Pervasive Systems: Examples
Sensor Networks
 Consists of spatially distributed autonomous sensors to
cooperatively monitor physical or environmental conditions,
such as temperature, sound, vibration, pressure, motion or
pollutants, etc.
 The nodes to which sensors are attached are:
 Many (10s-1000s)
 Simple (i.e., hardly any memory, CPU power, or
communication facilities)
 Often battery-powered

38
 Related Concepts : Parallel Systems
 Parallel Computing is a form of computation in which many
calculations are carried out simultaneously, operating on the
principle that large problems can often be divided into smaller
ones, which are then solved concurrently (in parallel).
 In parallel computing, all processors have access to a shared
memory. Shared memory can be used to exchange information
between processors.
 In distributed computing, each processor has its own private
memory (distributed memory). Information is exchanged by
passing messages between the processors.

39
 Related Concepts : Centralized System
 Centralized System Characteristics
 One component with non-autonomous parts.
 Component shared by users all the time.
 All resources accessible.
 Software runs in a single process.
 Single point of control.
 Single point of failure.
 Distributed System Characteristics
 Multiple autonomous components.
 Components are not shared by all users.
 Resources may not be accessible.
 Software runs in concurrently on different processors.
 Multiple points of control.
 Multiple points of failure.
40
41
?