Distributed System

• Distributed system is collection of

loosely coupled processors
interconnected by a
communications network

• A system in which hardware or

software components located at
networked computers
communicate and coordinate their
actions only by message passing”

• Processors variously called nodes,

computers, machines, hosts
 Motivation
 Reasons for distributed systems:
 Resource sharing
Sharing printing and files at remote sites
Processing information in a distributed database
Using remote specialized hardware devices

Using specialized software at remote site

 Computation speedup
 load sharing by moving jobs to lightly loaded sites
 Parallel processing
 Reliability – detect and recover from site failure, function
transfer, reintegrate failed site
 Communication – message passing
File transfer, login, mail, and RPC
 Type of Distributed System

Different types of distributed

system
Cluster Computing
Grid Computing
Cloud Computing
Peer-to-Peer

Distributed Storage System

 Cluster Computing Systems
 Cluster computing : Collection of
Processing nodes interconnected
by LAN and Switches.
 Works as integrated collection of
resources (Processors + storage)

 Scalable and Resilient (all nodes are

replaceable).
 Resource sharing among different
users.
 High latency, low bandwidth and
administrative complexity.

5
 Grid Computing Systems
 Grid computing : A grid
computing is a distributed
infrastructure and collections of
resources such as databases,
storage servers, high-speed
networks, supercomputers and
computer clusters available for
solving large-scale modern
sciences problems.

 Grid is an evolution of distributed computing

 The objective of the grid  Dynamic
technology is to share the vast  Geographically independent,
amount of resources that are  Built around standards (Open Grid
Forum),
owned by distinctive individuals
 Internet backbone
through secure and reliable
communication
6
 Cloud Computing Systems
 Cloud computing: provide scalable,
virtual, shared and on-demand
computing resources across the
Internet to small, medium and
large enterprises.

 Security consideration:
 confidentiality,
IaaS (shared storage and computing resources)  integrity (of data) and
PaaS (platform for testing any application)  availability
SaaS (software and applications)  Trusted Clod Computing.

7
 Amazon Web Services
• Elastic Compute Cloud (EC2)
–Rent computing resources pay-as-you-go
–Basic unit of accounting = instance/hour
–Additional costs for bandwidth
• Simple Storage Service (S3)
–Persistent storage
–Charge by the GB/month ($0.15/Month)
–Additional costs for bandwidth (sending and
receiving data)
 Peer-to-Peer (P2P) Network
 A Peer-to-Peer (P2P) network is created when two or
more PCs are connected and share resources without
going through a separate server computer
 P2P networks are de-centralised distributed systems,
which enable applications such as
 file-sharing,
 instant messaging,
 online multiuser
 gaming and
 content distribution over public networks
 Example of P2P: BitTorrent
 Distributed Storage System
 “Storing data on multitude of standard servers, which
behave as one storage system although data is
distributed between these servers”.
 Distributed storage systems such as NFS (Network
File System) provide users with a unified view of
data stored on different file systems and computers
which may be on the same or different networks.
 Examples:
 Google BigTable
 GFS (Google File System)
 HDFS (Hadoop Distributed File System)
 Types of Distributed Computing
Operating Systems

 Two types of distributed computing operating systems:

 Network Operating Systems

 Distributed Operating Systems

 Network-Operating Systems
 A network operating system (NOS) is a computer operating system that is
designed primarily to support workstation, personal computer, and
terminal that are connected on a local area network (LAN).
 A network operating system provides printer sharing, common file system
and database sharing, application sharing, and the ability to manage a
network name directory, security, and other housekeeping aspects of a
network.
 Users are aware of multiplicity of machines, not transparent and more
difficult to use
 Access to resources of various machines is done explicitly by:
 Remote logging into the appropriate remote machine (telnet, ssh)
 Remote Desktop
 Transferring data from remote machines to local machines, via the File
Transfer Protocol (FTP) mechanism
 Requires explicit FTP commands: get, put…
Overview of NOS Characteristics
• Network operating systems (NOSs)
distribute their functions over a
number of networked computers.
• It then adds functions that allow
access to shared resources by a
number of users concurrently.
• NOS computers take on specialized
roles to accomplish concurrent access
to shared resources.
• Client systems contain specialized
software that allows them to request
shared resources that are controlled
by server systems responding to a
client request.
Multiuser, Multitasking, and
Multiprocessor Systems
• A NOS server is a multitasking system.
Internally, the OS must be capable of
executing multiple tasks or processes
at the same time.
• Some systems are equipped with more
than one processor, called
multiprocessing systems.
• They are capable of executing multiple
tasks in parallel by assigning each task
to a different processor.
• The aggregate amount of work that
the server can perform in a given time
is greatly enhanced in multiprocessor
systems.
Choosing a NOS
• The main features to
consider when selecting a
NOS include:
– Performance
– Management and
monitoring tools
– Security
– Scalability
– Robustness/fault
tolerance
Types of NOS
• It is important to know the basics
about popular NOS families.
• Many networks now include more
than one server type, and knowing
how to get these diverse systems
to interoperate is an important skill
for a network administrator.
Windows NT and Windows 2000
Windows Terminology
• Windows server-based
networks that run Windows
NT Server or Windows 2000
Server are based on the
concept of the domain.
• A domain is a group of
computers and users that
serves as a boundary of
administrative authority.
• Windows NT domains and
Windows 2000 domains,
although similar in
function, interact with one
another differently.
Windows NT 4.0
• The Domain Structure of
Windows NT was entirely
different from the Domain
Structure in Windows 2000.
• Instead of Active Directory,
Windows NT provides an
administrative tool called
the User Manager for
Domains.
• It is accessed from the
domain controller and is
used to create, manage,
and remove domain user
accounts.
 Windows NT 4.0

• Each NT domain requires one Primary Domain Controller

(PDC).
• This is a "master" server that contains the Security Accounts
Management Database (SAM).
• A domain can also have one or more Backup Domain
Controllers (BDCs), each of which contains a read-only copy
of the SAM.
• The SAM is what controls the authentication process when a
user logs onto the domain.
 Windows 2000 Operating System
• Administrative tasks in Windows
2000 use a common framework,
the Microsoft Management
Console (MMC).
• This tool uses snap-ins, which are
modules that contain the tools for
specific administrative functions.
• Users and groups are created and
managed with the Active Directory
Users (ADUs) and Computers
• An organizational unit (OU) is a MMC snap-in.
container within a Microsoft Active
Directory domain which can hold • Administrative authority over each
users, groups and computers. OU (Organization Unit) can be
• It is the smallest unit to which an delegated to a user or group.
administrator can assign Group
Policy settings or account
permissions
Windows 2000's Family of Operating
Systems
• The Windows 2000 family
of operating systems
includes:
– Windows 2000
Professional
– Windows 2000 Server
– Windows 2000
Advanced Server
• The specific needs of the
network will determine
the best version of
Windows 2000 for the
installation.
Linux
 History of Linux
• Linux is an operating system similar to UNIX. It runs on
many different computers and was first released in
1991.
• Linux is portable, which means versions can be found
running on name brand or clone PCs.
• Linux offers many features adopted from other versions
of UNIX.
What is UNIX?
• The UNIX NOS was developed in
1969, and it has evolved into
many varieties.
• The source code is opened, that
is, available at no cost to anyone
who wants to modify it.
• It is written in C programming
language so businesses,
academic institutions, and even
individuals can develop their
own versions.
• There are hundreds of different
versions of UNIX.
 Linux Operating System
• Linux is sometimes referred to as "UNIX Lite", and it is
designed to run on Intel-compatible PCs.
• However, Linux will run on other machines as well.
• Linux brings the advantages of UNIX to home and small
business computers.
• The following are a few of the most popular types:
– Red Hat Linux
– Linux Mandrake
– Caldera eDesktop and eServer
– Debian GNU/Linux
– Corel Linux
– Turbo Linux
Linux Clients
• Windows clients can access
Linux servers without client
software if the UNIX servers
run Samba, which is a
program that uses the Server
Message Block (SMB)
application layer protocol.
• Windows computers use SMB
for file access across the
network.
• Samba permits them to see
the Linux file system.
 Distributed-Operating Systems

 Distributed operating system is a model where distributed applications

are running on multiple computer linked by communications.
 A distributed operating system is an extension of the network operating
system that supports higher levels of communication and integration of
the machines on the network.
 Users not aware of multiplicity of machines: access to remote
resources similar to access to local resources
 Distributed-Operating Systems
Characteristics of Distributed Systems
 Data Migration – transfer data by transferring entire file, or
transferring only those portions of the file necessary for the
immediate task
 Old version of Andrew file system (AFS) moves entire file
to local site (automated FTP)
The Andrew file system is a distributed file system
which uses a set of trusted servers to present a
homogeneous, location-transparent file name space to
all client workstation
 NFS only moves parts that are needed
The NFS is a client/server application that lets a
computer user view and optionally store and update file
on remote computer. The system need NFS client and
NFS server.
 Distributed-Operating Systems
(Cont.)
 Computation Migration – transfer the
computation, rather than the data, across
the system
 If it takes longer to transfer the data than
it is to execute the command, then
migrate operation
Database queries

 Process Migration – execute an entire

process, or parts of it, at different sites.
 Distributed-Operating Systems
(Cont.)
 Load balancing – distribute processes across network
to even the workload
 Computation speedup – subprocesses can run
concurrently on different sites
 Reliability – Long term data preservation and backup
(replication) at different locations.
 Scalability – Addition of more resources to increase
performance or availability.
 Data access – run process remotely, rather than
transfer all data locally.
 Economy - Sharing of resources by many entities to
help reduce the cost of ownership.
 Distributed-Operating Systems (Cont.)
 Heterogeneity
 Various entities in the system must be able to
interoperate with one another, despite differences in
hardware architectures, operating systems,
communication, protocols, programming languages,
software interfaces, security models, and data formats.
 Transparency
 The entire system should appear as a single unit and the
complexity and interactions between the components
should be typically hidden from the end user.
 Fault tolerance and failure management
 Failure of one or more components should not bring down
the entire system, and should be isolated.
Transparency in a Distributed System