Distributed System

Chapter One

2704
867 BOS

849 PVD
ORD 187
740 144
1846 621 JFK
184 1258
802
SFO BWI
1391
1464
337 1090
DFW 946
LAX 1235
1121
MIA
10/16/2024 2342 By Lewyehu Y. 1
1. Introduction and Definition
 Before-1980
Computer were large and expensive
very slow (a few thousand instructions per second)
 not connected among themselves
All systems were centralized systems
 after the mid-80s: two major developments
cheap and powerful microprocessor-based computers appeared
High speed computer networks(LANs , WANs)

Then came to the Distributed System

10/16/2024 By Lewyehu Y. 2
1.1 DEFINITION OF A DISTRIBUTED SYSTEM
 A distributed system is a collection of independent computers
that appears to its users as a single coherent system.

 The distributed system “looks like” a single computer rather

than a collection of separate computers.

 Distributed System is a collection of independent computers

(processing elements) which are geographically (spatially)
distributed and do not share memory or clock and functions
like a single system for the users.

10/16/2024 By Lewyehu Y. 3
 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

10/16/2024 By Lewyehu Y. 4
 Other Definitions
◊ A distributed system is a system designed to support the
development of applications and services which can exploit a
physical architecture consisting of multiple, autonomous
processing elements that do not share primary memory but
cooperate by sending asynchronous messages over a
communication network (Blair & Stefani).

◊ Examples of Distributed systems: Internet, E-mail, E-commerce,

Banking, Online gaming, Social networks, Sensor networks, Mobile
and Pervasive Computing.

10/16/2024 By Lewyehu Y. 5
 Organization of a Distributed System
 to support heterogeneous computers and networks and to
provide a single-system view, a distributed system is often
organized by means of a layer of software called middleware
that extends over multiple machines

Figure 1.1 a distributed system organized as middleware; note that the

middleware layer extends over multiple machines, and offers each application
the same interface
10/16/2024 By Lewyehu Y. 6
 A distributed system organized as middleware.

 The middleware layer runs on all machines, and offers a

uniform interface to the system.

 Middleware: a software layer that provides a programming

abstraction to mask the heterogeneity of the underlying
platforms (networks, languages, H/W, …)

 In distributed computing a program is split up into parts that

run simultaneously on multiple computers communicating
over a network.

10/16/2024 By Lewyehu Y. 7
 Role of Middleware (MW)
 Middleware allows independent computers to work together
closely.

 MW also supports seamless access to remote services.

 middleware offers an infrastructure that enables application
processes to communicate with each other

10/16/2024 By Lewyehu Y. 8
Middleware Examples

 CORBA (Common Object Request Broker Architecture)

 DCOM (Distributed Component Object Management) –
being replaced by .net

 RPC (Remote Procedure Call)

 RMI (Remote Method Invocation)
 SOAP (Simple Object Access Protocol)

10/16/2024 By Lewyehu Y. 9
 All of the previous examples support communication across a
network:

 They provide protocols that allow a program running on one kind

of computer, using one kind of operating system, to call a program
running on another computer with a different operating system

 The communicating programs must be running the same

middleware.

10/16/2024 By Lewyehu Y. 10
 Advantages of Distributed system?
 Economically: a collection of microprocessors offer a better
price/performance than mainframes

 Speed: a distributed system may have a total computing power

than a mainframe. 10,000 CPU chips, each running

at 50 MIPS(million instruction per second).

 Reliability: If one machine crashes, the system as a whole can still

survive.

 Incremental growth: Computing power can be added in

small increments. Modular expandability

10/16/2024 By Lewyehu Y. 11
  Resource and Data Sharing
• printers, databases, multimedia servers, ...
 Scalability, Extensibility
• the system grows with demand (e.g., extra servers)
 Performance
• huge power (CPU, memory, ...) available
 Inherent distribution, communication
• organizational distribution, e-mail, video

10/16/2024 By Lewyehu Y. 12
 Disadvantage of Distributed system
Security: Easy access also applies to secret data
Privacy: unwanted communication such as spam
Partial failure: we often do not know where the error is
Software: difficult to develop software for distributed
systems

Network: loss transmissions

10/16/2024 By Lewyehu Y. 13
 Characteristics of Distributed Systems
 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

 Multiple autonomous components

 Software runs a concurrent processes in different processors
 Multiple point of controllers
 Multiple point of failure
 Supported by middleware
10/16/2024 By Lewyehu Y. 14
1.2 . Goals of a distributed system(Why Distributed Systems?)
 There are important goals that should be met to make building a
distributed system worth the effort.

 Resources Accessible : A distributed system should make

resources easily accessible.

 Distribution Transparency : It should reasonably hide the fact

that resources are distributed across a network.

 Openness: It should be open.

 Scalability : It should be scalable and it should be

 concurrency
10/16/2024 By Lewyehu Y. 15
  Resource sharing (Making Resources Accessible)

 The main goal of a distributed system is to make it easy for the users
(and applications) to access remote resources, and to share them in a
controlled and efficient way.

 a distributed system should also make it easier for users to exchange

information’s.

 Resources can be just about anything, but typical examples include

things like printers, computers, storage facilities, data, files, Web
pages, and networks, to name just a few.

 There are many reasons for wanting to share resources. One obvious
reason is that of economics.
10/16/2024 By Lewyehu Y. 16
 Connecting users and resources also makes it easier to
collaborate and exchange information, as is clearly illustrated by
the success of the Internet with its simple protocols for
exchanging files, mail, documents, audio, and video.

 Distribution Transparency

 An important goal of a distributed system is to hide the fact that

its processes and resources are physically distributed across
multiple computers.

10/16/2024 By Lewyehu Y. 17
 Transparency means that any distributed system should hide its
distributed nature from its users appearing and functioning as a
normal centralised system.
 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.
 Types of Transparency
There are many types of transparency:
 Access transparency: resources are accessed in a single, uniform
way (regardless of whether they are local or remote). Hide
differences in data representation and how resources are accessed.

10/16/2024 By Lewyehu Y. 18
 Location transparency: users should not be aware of where a
resource is physically located. That is hide where resources are
located.
 Concurrency transparency: multiple users may compete for and
share a single resource: this should not be apparent to them, hide
that resource may be shared by several competitive users.
 Replication transparency: even if a resource is replicated, it should
appear to the user as a single resource (without knowledge of the
replicas). Hide that resources are replicated.
 Failure transparency: always try to hide any faults. …and more:
mobility (migration/relocation), performance, scaling, persistence,
security , hides failure and recovery of resource
 Relocation : Hide that resource may be moved to another location
while in use.
 Migration: Hide that resources may moved to another location.
10/16/2024 By Lewyehu Y. 19
  Openness
 Openness is the property of a distributed system such that each
subsystem is continually open to interaction with other systems.

 As a result the key software interfaces of the components of such a

system are made publicly available.
 a distributed system should be open
 we need well-defined interfaces
 interoperability
 components of different origin can communicate
 portability
 components work on different platforms

10/16/2024 By Lewyehu Y. 20
 Another goal of an open distributed system is that it should be
flexible and extensible; easy to configure the system out of
different components; easy to add new components, replace
existing ones.
 An Open Distributed System is a system that offers services
according to standard rules that describe the syntax and semantics
of those services; e.g., protocols in networks
 In distributed systems, such services are often specified through
interfaces often described using an Interface Definition Language
(IDL)
 Specify only syntax: the names of the functions, types of
parameters, return values,
10/16/2024
possible exceptions,….
By Lewyehu Y. 21
 Interface Definition/Description Languages (IDL): used to
describe the interfaces between software components, usually
in a distributed system.

Definitions are language & machine independent

Support communication between systems using different
OS/programming languages; e.g. a C++ program running
on Windows communicates with a Java program running
on UNIX

Communication is usually RPC-based.

10/16/2024 By Lewyehu Y. 22
  Scalability
• A system is described scalable if it will remain effective, without
disproportional performance loss, when there is an increase in the
number of resources, the number of users, or the amount of input
data.

• Scalability in terms of: load, geographical distribution, number of

different organisations…

• Scalability of a system can be measured along at least three

different dimensions (Neuman, 1994).

10/16/2024 By Lewyehu Y. 23
 A distributed system should be scalable
 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.

 But system that is scalable in one or more of these dimensions

often exhibits some loss of performance as the system scales up.

10/16/2024 By Lewyehu Y. 24
 Scalability Problems
 When a system needs to scale, very different types of problems need to be solved.

Let us first consider scaling with respect to size.

 If more users or resources need to be supported, we are often confronted with the

limitations of centralized services, data, and algorithms.

 For example, many services are centralized in the sense that they are

implemented by means of only a single server running on a specific machine in

the distributed system.

 The problem with this scheme is obvious: the server can become a bottleneck as

the number of users and applications grows.

 Even if we have virtually unlimited processing and storage capacity,

communication with that server will eventually prohibit further growth.

10/16/2024 By Lewyehu Y. 25
 Using only a single server is sometimes unavoidable.
 Imagine that we have a service for managing highly confidential
information such as medical records, bank accounts and so on.

Examples of scalability limitations

10/16/2024 By Lewyehu Y. 26
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)

 three possible solutions and techniques to improve scalability :

hiding communication latencies, distribution, and replication

10/16/2024 By Lewyehu Y. 27
 a. Hide Communication Latencies
try to avoid waiting for responses to remote service requests
Have separate handler for incoming response
i.e., construct requesting applications that use only asynchronous
communication instead of synchronous communication; when a
reply arrives the application is interrupted.

 Do some work at client side or e.g. do other useful work at the

requester's side

 Problem: not every application fits this model

10/16/2024 By Lewyehu Y. 28
 (a) a server checking the correctness of field entries
(b) a client doing the job

• e.g., checking the completeness of mandatory fields

10/16/2024 By Lewyehu Y. 29
b. distribution
 Another important scaling technique is distribution.

 Distribution involves taking a component, splitting it into

smaller parts, and subsequently spreading those parts across the
system .

 An excellent example of distribution is the Internet Domain

Name System (DNS).

 The DNS name space is hierarchically organized into a tree of

domains, which are divided into non-overlapping zones, as
shown in Fig. 1-5. The names in each zone are handled by a
single name server.
10/16/2024 By Lewyehu Y. 30
 an example of dividing the DNS name space into zones

10/16/2024 By Lewyehu Y. 31
c. Replication
 replicate components across a distributed system to increase
availability and for load balancing, leading to better performance

 Replicated file servers and databases

 decided by the owner of a resource
 caching (a special form of replication) also reduces
communication latency; decided by the user

 Web caches (in browsers and proxies)

 File caching (at server and client)
 but, caching and replication may lead to consistency problems

10/16/2024 By Lewyehu Y. 32
  Concurrency
 There is a possibility that several clients will attempt to access a
shared resource at the same time.

 Any object that represents a shared resource in a distributed

system must be responsible for ensuring that operates correctly
in a concurrent environment.
• Several clients may attempt to access a shared resource at the same
time.
– E.g., eBay/Amazon bids
• Requires proper synchronisation to make sure that data remains
consistent.

10/16/2024 By Lewyehu Y. 33
 Pitfalls when Developing Distributed Systems
(Assumptions which are usually wrong for DS)
Peter Deutsch, then at Sun Microsystems, formulated these mistakes
as the following false assumptions that everyone makes when
developing a distributed application for the first time:
1. The network is reliable
2. Latency is zero
3. Bandwidth is infinite
4. The network is secure
5. Topology doesn’t change
6. There is one administrator
7. Transport cost is zero
8. The network is homogeneous

10/16/2024 By Lewyehu Y. 34
 1.3 TYPES OF DISTRIBUTED SYSTEMS

 Three types: distributed computing systems, distributed

information systems, and distributed embedded systems

1 . Distributed Computing Systems

 An important class of distributed systems is the one used for high-

performance computing tasks.

Distributed Computing Systems: can be classified as

a. Cluster Computing

b. Grid Computing

c. Cloud Computing)
10/16/2024 By Lewyehu Y. 35
 a. Cluster Computing
 Cluster computing involves a group of interconnected computers (or
nodes) that work together as a single system to perform tasks.

 These nodes are usually located close to each other, often within the
same physical location.

 A collection of similar processors (PCs, workstations) running the

same operating system, connected by a high-speed LAN.

 Cluster computing is designed to provide high availability, load

balancing, and parallel processing for computational tasks.

10/16/2024 By Lewyehu Y. 36
 A characteristic feature of cluster computing is its
homogeneity. In most cases, the computers in a cluster are
largely the same.
 They all have the same operating system, and are all connected
through the same network.

10/16/2024 By Lewyehu Y. 37
Cluster Types & Uses
 High Performance Clusters (HPC) : Run large parallel programs as like Scientific,
military, engineering apps; e.g., weather modelling, Web Server Clusters.
 Load Balancing Clusters (LBC):
 Load Balancing Clusters are designed to distribute incoming requests across multiple
servers to ensure no single server becomes overwhelmed.
 Front end processor distributes incoming requests, server farms (e.g., at banks or popular
commercial and financial web sites).
 Example: Banks use load balancers to direct customer requests across various servers,
maintaining performance during busy periods.
 High Availability Clusters (HA) : provide redundancy – back up systems, May be more
fault tolerant than large scale mainframe systems.
 High Availability Clusters are designed to ensure that systems remain operational and
available, even in the event of hardware or software failures.
 Telecommunications

10/16/2024 By Lewyehu Y. 38
 b. Grid Computing
 Grid computing is a distributed computing model that connects
multiple heterogeneous computing resources across different
locations to work together on complex problems.

 Grid computing systems have a high degree of heterogeneity: no

assumptions are made concerning hardware, operating systems,
networks, administrative domains, security policies, etc.

 Similar to clusters but processors are more loosely coupled, tend to

be heterogeneous, and are not all in a central location.

 Problems are broken up into parts and distributed across multiple

computers in the grid.
10/16/2024 By Lewyehu Y. 39
How it works? (Resource Sharing, Job Distribution, Middleware,
Scalability)

The architecture consists of four layers.

 fabric layer: provides direct interfaces to local resources at a
specific site.
 connectivity layer: consists of communication protocols for
supporting grid transactions that span the usage of multiple
resources.

 resource layer: is responsible for managing a single resource.

- It uses the functions provided by the connectivity layer and calls
directly the interfaces made available by the fabric layer.

10/16/2024 By Lewyehu Y. 40
• The collective layer: It deals with handling access to multiple
resources and typically consists of services for resource discovery,
allocation and scheduling of tasks onto multiple resources, data
replication, and so on.

• The application layer : consists of the applications that operate

within a virtual organization and which make use of the grid
computing environment.

10/16/2024 By Lewyehu Y. 41
10/16/2024 By Lewyehu Y. 42
 c. Cloud Computing
 Cloud computing is the delivery of computing services—including
servers, storage, databases, networking, software, analytics, etc…
over the internet (the cloud).

 Service Models:

 Infrastructure as a Service (IaaS): Provides virtualized computing

resources over the internet.

 Platform as a Service (PaaS): Offers hardware and software tools over

the internet, allowing developers to build, test, and deploy
applications without managing the underlying infrastructure

10/16/2024 By Lewyehu Y. 43
 Software as a Service (SaaS): Delivers software applications
over the internet on a subscription basis, eliminating the need
for installation and maintenance.

10/16/2024 By Lewyehu Y. 44
2. Distributed Information Systems
 Distributed information systems manage and process data across
multiple nodes, enabling data sharing and collaboration among
users in different locations.

 These systems enable the sharing of data and resources over a

network, allowing multiple users and applications to access and
manipulate data in a coordinated manner.

 distribute information across several servers.

 remote processes called clients access the servers to manipulate

the information.

10/16/2024 By Lewyehu Y. 45
2.1.Transaction processing systems
 Transaction processing systems are specialized software systems
designed to handle transactions in a consistent, reliable, and
efficient manner.
 They are critical in environments where a high volume of
transactions needs to be processed, such as banking,
e-commerce, and inventory management.
 Highly Structured Approach: TPS typically use a client-server
model where the client (e.g., ATM) sends requests to a server
(e.g., bank database) to perform operations such as withdrawal,
deposit, or balance inquiry.

10/16/2024 By Lewyehu Y. 46
 In practice, operations on a database are usually carried out in
the form of transactions.

 Programming using transactions requires special primitives

that must either be supplied by the underlying distributed
system or by the language runtime system.

 Typical examples of transaction primitives are shown in Fig.

1- 8. The exact list of primitives depends on what kinds of
objects are being used in the transaction.

10/16/2024 By Lewyehu Y. 47
10/16/2024 By Lewyehu Y. 48
 The Transaction Model
 the model for transactions comes from the world of business
 a supplier and a retailer negotiate on
 price
 delivery date
 Quality etc.
 until the deal is concluded they can continue negotiating
or one of them can terminate
 but once they have reached an agreement they are bound
by law to carry out their part of the deal
 transactions between processes is similar with this
scenario

10/16/2024 By Lewyehu Y. 49
  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

10/16/2024 By Lewyehu Y. 50
 (a) transaction to reserve three flights commits
(b) transaction aborts when third flight is unavailable

10/16/2024 By Lewyehu Y. 51
  properties of transactions, often referred to as ACID
1.Atomic: all statement in the transaction either completed
successfully or they were all rolled back.
- A transaction either happens completely or not at all;
intermediate states are not seen by other processes
1.Consistent: the transaction does not violate system invariants.
- The transaction is logically consistent state ; e.g., in an internal
transfer in a bank, the amount of money in the bank must be the
same as it was before the transfer (the law of conservation of
money);
1.Isolated or Serializable: concurrent transactions do not
interfere with each other; if two or more transactions are
running at the same time, the final result must look as though
all transactions run sequentially in some order
2.Durable: once a transaction commits, the changes are
permanent

10/16/2024 By Lewyehu Y. 52
 Classification of Transactions
 a transaction could be flat, nested or distributed
 Flat Transaction
 consists of a series of operations that satisfy the ACID
properties
 simple and widely used but with some limitations
 some transactions may take too much time
 Nested Transaction
 constructed from a number of sub-transactions; it is logically
decomposed into a hierarchy of sub-transactions
 that run in parallel, on different machines; to gain performance
 each may also execute one or more sub-transactions

10/16/2024 By Lewyehu Y. 53
 Nested transactions are important in distributed systems
 they provide a natural way of distributing a transaction across multiple
machines.
 They follow a logical division of the work of the original transaction and
managed by the software component.
 This component was called a transaction processing monitor or TP monitor
for short.
 Its main task was to allow an application to access multiple
server/databases by offering it a transactional programming model, as
shown in Fig. 1-10.

10/16/2024 By Lewyehu Y. 54
  Distributed Transaction
 a flat transaction that operates on data that are distributed
across multiple machines

10/16/2024 By Lewyehu Y. 55
 2.2. Enterprise Application Integration

 Less structured than transaction-based systems.

 Enterprise applications are things like HR data, inventory
programs, etc.,
 May use different OSs, different DBs but need to interoperate
sometimes.
 Communication mechanisms to support this include CORBA,
Remote Procedure Call (RPC) and Remote Method Invocation
(RMI).

10/16/2024 By Lewyehu Y. 56
10/16/2024 By Lewyehu Y. 57
 3. Distributed Pervasive Systems
 A pervasive system, also known as ubiquitous computing, refers to a
computing environment where technology is seamlessly integrated into the
background of everyday life.
 The goal of pervasive systems is to enhance the user's experience by
making technology available at any time and in any place, often without
the user being explicitly aware of it.
• The first two types of systems are characterized by their stability: nodes
and network connections are more or less fixed but this type of system is
likely to incorporate small, battery-powered, mobile devices.
 Home systems
 Electronic health care systems – patient monitoring
 Sensor networks – data collection, surveillance
10/16/2024 By Lewyehu Y. 58
 Hence, Distributed systems are everywhere - Internet, intranet,
wireless networks.

 Resource sharing is the main motivating factor for constructing

distributed systems.

 The construction of distributed systems produces many

challenges like secure communication over public networks.

10/16/2024 By Lewyehu Y. 59
 In order to design a good distributed system, there are six key
design goals:
 Concurrency
 Scalability
 Openness
 Fault Tolerance
 Privacy & Authentication and
 Transparency.

10/16/2024 By Lewyehu Y. 60
10/16/2024 By Lewyehu Y. 61