Unit: 1 Introduction to

Distributed System
1) Define Distributed Operating System and Explain Goals of Distributed System.
 A distributed system is a collection of independent computers that appear to the users of the
system as a single computer.
 Using high performance computers connected by equally high speed communication links, it is
possible to build a single system consisting of multiple computer and using it as a single
consolidated system.
 In such a system, multiple resources work together to deliver the required processing speed
and the operating system takes care of maintaining the system and overall maintenance.
 In distributed system computers are not independent but interconnected by a high speed
network.
 It means that many computers, be they workstation or desktop systems linked together, can
do work of a high performance supercomputer.
 If user thinks that the entire interlinked system is a single and unified computer, there has to
be software envelope that joins multiple operating system together and offers a seamless
interface to user.

The following are the main goals of distributed systems:

 The relative simplicity of the software - each processor has a dedicated function.
 Incremental growth - if we need 10 percent more computing power, we just add 10 percent
more processors.
 Reliability and availability - a few parts of the system can be down without disturbing people
using the other parts.
 Openness: Offer services according to standard rules that describe the syntax and semantics of
those services.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 1

 Unit: 1 Introduction to
Distributed System
2) Explain Advantage of Distributed system over centralized system.
Economics
 A quarter century ago, according to Grosch’s law: the computing power of a CPU is proportional
to the square of its price.
 By paying twice as much you could get four times the performance.
 This observation fit the mainframe technology of its time.
 With microprocessor technology, Grosch’s law no longer holds.
 For a few hundred dollars you can get the CPU chip that can execute more instructions per
second than one of the largest 1980s mainframe.
Speed
 A collection of microprocessors cannot only give a better price/performance ratio than a single
mainframe, but also give an absolute performance that no mainframe can achieve at any price.
 For example, with current technology it is possible to build a system from 10,000 modern CPU
chips, each of which runs at 50 MIPS (Millions of Instructions Per Second), for a total
performance of 500,000 MIPS.
 For a single processor (i.e., CPU) to achieve this, it would have to execute an instruction in 0.002
nsec (2 picosecond).
Inherent Distribution
 Many institutions have applications which are more suitable for distributed computation.
 Assume that there is large company buying and selling goods from different countries.
 Its offices situated in those countries are geographically diverse.
 If a company wishes unified computing system, it should implement a distributed computing
system.
 Consider the global employee database of such a multinational company.
 Branch offices would create local database and then link them for global viewing.
Reliability
 By distributing the workload over many machines, a single chip failure will bring down at most
one machine, leaving the rest intact.
 Ideally, if 5 percent of the machines are down at any moment, the system should be able to
continue to work with a 5 percent loss in performance.
 For critical applications, such as control of nuclear reactors or aircraft, using a distributed
system to achieve high reliability may be the dominant consideration.
Incremental Growth
 A company will buy a mainframe with the intention of doing all its work on it.
 If the company expands and the workload grows, at a certain point the mainframe will no
longer be adequate.
 The only solutions are either to replace the mainframe with a larger one (if it exists) or to add
a second mainframe.
 Both of these can cause damage on the company's operations.
 In contrast, with a distributed system, it may be possible simply to add more processors to the
system, thus allowing it to expand gradually as the need arises.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 2

 Unit: 1 Introduction to
Distributed System
3) Explain Advantage of Distributed system over Independent PCs.
Data sharing
 Many users need to share data.
 For example, airline reservation clerks need access to the master data base of flights and
existing reservations.
 Giving each clerk his own private copy of the entire data base would not work, since nobody
would know which seats the other clerks had already sold.
 Shared data are absolutely essential for this and many other applications, so the machines must
be interconnected.
Resource sharing
 Expensive peripherals, such as color laser printers, phototypesetters, and massive archival
storage devices (e.g., optical jukeboxes), can also be shared.
Communication
 For many people, electronic mail has numerous attractions over paper mail, telephone, and
FAX.
 It is much faster than paper mail, does not require both parties to be available at the same time
as does the telephone, and unlike FAX, produces documents that can be edited, rearranged,
stored in the computer, and manipulated with text processing programs.
Flexibility
 A distributed system is more flexible than giving each user an isolated personal computer.
 One way is to give each person a personal computer and connect them all with a LAN, this is
not the only possibility.
 Another one is to have a mixture of personal and shared computers, perhaps of different sizes,
and let jobs run on the most appropriate one, rather than always on the owner's machine.
 In this way, the workload can be spread over the computers more effectively, and the loss of a
few machines may be compensated for by letting people run their jobs elsewhere.
4) Classify the Network operating system and Distributed operating system.
Network Operating system Distributed Operating system
A network operating system is made up of A distributed operating system is an ordinary
software and associated protocols that allow a centralized operating system but runs on
set of computer network to be used together. multiple independent CPUs.
Environment users are aware of multiplicity of Environment users are not aware of
machines. multiplicity of machines.
Control over file placement is done manually It can be done automatically by the system
by the user. itself.
Performance is badly affected if certain part of It is more reliable or fault tolerant i.e.
the hardware starts malfunctioning. distributed operating system performs even if
certain part of the hardware starts
malfunctioning.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 3

 Unit: 1 Introduction to
Distributed System

Remote resources are accessed by either Users access remote resources in the same
logging into the desired remote machine or manner as they access local resources.
transferring data from the remote machine to
user's own machines.
5) Classify Distributed operating system.

Figure: Classification of distributed system

Hardware Concepts
 Even though all distributed systems consist of multiple CPUs, there are several different ways
the hardware can be organized, especially in terms of how they are interconnected and how
they communicate.
 Various classification schemes for multiple CPU computer systems have been proposed over
the years.
 According to Flynn, classification can be done based on, the number of instruction streams and
number of data streams.
SISD
 A computer with a single instruction stream and a single data stream is called SISD.
 All traditional uniprocessor computers (i.e., those having only one CPU) fall in this category,
from personal computers to large mainframes.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 4

 Unit: 1 Introduction to
Distributed System
SIMD
 The next category is SIMD, single instruction stream, multiple data stream.
 This type refers to array processors with one instruction unit that fetches an instruction, and
then commands many data units to carry it out in parallel, each with its own data.
 These machines are useful for computations that repeat the same calculation on many sets of
data, for example, adding up all the elements of 64 independent vectors.
 Some supercomputers are SIMD.

MISD
 The next category is MISD, multiple instruction stream, single data stream, no known
computers fit this model.

MIMD
 Finally, comes MIMD, which essentially means a group of independent computers, each with
its own program counter, program, and data.
 All distributed systems are MIMD.
 We divide all MIMD computers into two groups: those that have shared memory, usually called
multiprocessors, and those that do not, sometimes called multicomputer.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 5

 Unit: 1 Introduction to
Distributed System

Software Concepts
 The image that a system presents to its users, and how they think about the system, is largely
determined by the operating system software, not the hardware.
 There are basic two types of operating system namely (1) Tightly coupled operating system and
(2) loosely couple operating system; for multiprocessor and multicomputer.
 Loosely-coupled software allows machines and users of a distributed system to be
fundamentally independent of one another.
 Consider a group of personal computers, each of which has its own CPU, its own memory, its
own hard disk, and its own operating system, but which share some resources, such as laser
printers and data bases, over a LAN.
 This system is loosely coupled, since the individual machines are clearly distinguishable, each
with its own job to do.
 If the network should go down for some reason, the individual machines can still continue to
run to a considerable degree, although some functionality may be lost.
 For tightly coupled system consider a multiprocessor dedicated to running a single chess
program in parallel.
 Each CPU is assigned a board to evaluate, and it spends its time examining that board and all
the boards that can be generated from it.
 When the evaluation is finished, the CPU reports back the results and is given a new board to
work on.
 The software for this system, both the application program and the operating system required
to support it, is clearly much more tightly coupled than in our previous example.
Tightly Coupled System
 Distributed Operating system is a tightly coupled software on loosely coupled hardware.
 The goal of such a system is to create the illusion in the minds of the users that the entire
network of computers is a single timesharing system, rather than a collection of distinct
machines.
Characteristics:
o There is a single, global inter process communication mechanism so that any process
can talk to any other process.
Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 6
 Unit: 1 Introduction to
Distributed System
o There is a global protection scheme.
o Process management is also uniform everywhere.
o How processes are created, destroyed, started, and stopped does not vary from
machine to machine.
o The file system looks same from everywhere.
o Every file is visible at every location, subject to protection and security constraints.
o Each kernel has considerable control over its own local resources.
o For example, if swapping or paging is used, the kernel on each CPU is the logical place
to determine what to do swap or page.
6) Explain Distributed Computing System Models.
 Distributed Computing system models can be broadly classified into five categories.
Minicomputer Model
 It consists of a few minicomputers interconnected by a communication network.
 Each minicomputer usually has multiple users logged on to it simultaneously.
 For this several interactive terminals are connected to each minicomputer.
 Each user is logged onto one minicomputer, with remote access to other minicomputers.
 The network allows the user to access remote resources that are available on some machine
other than the user is currently logged.
 Minicomputer model can be used for resource sharing like information databases.
 ARPAnet is the example Distributed computing system based on the minicomputer model.

Figure: Minicomputer model

Workstation Model
 A workstation is a high-end microcomputer designed for technical or scientific applications.
 Each workstation is equipped with its own disk and is intended primarily to be used by one
person at a time (single-user computer), but they are commonly connected to a local area
network and run multi-user operating systems.
 Large number of workstations can lie idle resulting in a waste of CPU time.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 7

 Unit: 1 Introduction to
Distributed System
 Using a high speed LAN allows idle workstations to process jobs of users who are logged onto
other workstations, which have less computing powers.

Figure: Workstation Model

Workstation-server Model
 It is a network of personal workstations each with its own disk and local file system.
 Workstation with its own local disk is usually called a disk full workstation and a workstation
without a local disk is called a diskless workstation.
 When diskless workstations are used, file system to be used must be implemented by a disk
full workstation or by a minicomputer equipped with a disk for file storage.
 Other minicomputers are used for providing other services like database, print etc.

Figure: Workstation-Server Model

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 8
 Unit: 1 Introduction to
Distributed System
Processor-Pool Model
 When a user needs a very large amount of computing power for a short time, instead of
allocating a processor to each user, processors are pooled together to be shared by the users
as needed.
 Pool of processors contain a large number of microcomputers and mini-computers.
 Each processor in the pool has its own memory to load and run a system program or an
application program of the distributed computing system.
 Amoeba, Cambridge Distributed Computing system are examples of process-pool model based
Distributed systems.

Figure: Process-pool model

Hybrid Model
 Advantages of the workstation-server and processor-pool models are combined to build a
hybrid model.
 It is built on the workstation-server model with a pool of processors.
 Processors in the pool can be allocated dynamically for large computations, that cannot be
handled by the workstations, and require several computers running concurrently for efficient
execution.
 This model is more expensive to implement than the hybrid or the processor-pool model.
7) Explain Design Issues in Distributed Operating System.
The issues in designing a distributed system include:
Transparency
 Main goal of Distributed system is to make the existence of multiple computers invisible
(transparent) and provide single system image to user.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 9

 Unit: 1 Introduction to
Distributed System
 A transparency is some aspect of the distributed system that is hidden from the user
(programmer, system developer, application).
 A transparency is provided by including some set of mechanisms in the distributed system at a
layer below the interface such that a communication system appears to its users as a virtual
uniprocessor.
 While users hit search in google.com……. They never notice that their query goes through a
complex process before google shows them a result.
Access Transparency:
o It hides differences in data representation and how a resource is accessed by a user.
o Example, a distributed system may have a computer system that runs different
operating systems, each having their own file naming conventions.
o Differences in naming conventions as well as how files can be manipulated should be
hidden from the users and applications.
Location Transparency:
o Hides where exactly the resource is located physically.
o Example: by assigning logical names to resources like yahoo.com, one cannot get an
idea of the location of the web page’s main server.
Migration Transparency:
o Distributed system in which resources can be moved without affecting how the
resource can be accessed are said to provide migration transparency.
o It hides that the resource may move from one location to another.
Relocation Transparency:
o This transparency deals with the fact that resources can be relocated while it is being
accessed without the user who is using the application to know anything.
o Example: using a Wi-Fi system on laptops while moving from place to place without
getting disconnected.
Replication Transparency:
o Hides the fact that multiple copies of a resource could exist simultaneously.
o To hide replication, it is essential that the replicas have the same name.
o Consequently, as system that supports replication should also support location
transparency.
Concurrency Transparency:
o It hides the fact that the resource may be shared by several competitive users.
o Example: two independent users may each have stored their file on the same server
and may be accessing the same table in a shared database.
o In such cases, it is important that each user should not notice that the others are making
use of the same resource.
Failure Transparency:
o Hides failure and recovery of the resources.
o Example: a user cannot distinguish between a very slow or dead resource.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 10

 Unit: 1 Introduction to
Distributed System
o Same error message come when a server is down or when the network is overloaded
of when the connection from the client side is lost.
o So here, the user is unable to understand what has to be done, either the user should
wait for the network to clear up, or try again later when the server is working again.
Persistence Transparency:
o It hides if the resource is in memory or disk.
o Example: Object oriented database provides facilities for directly invoking methods on
storage objects.
o First the database server copies the object states from the disk i.e. main memory
performs the operation and writes the state back to the disk.
o The user does not know that the server is moving between primary and secondary
memory.
Reliability
 Reliability in terms of data means that data should be available without any errors.
 Distributed system where multiple processors are available and the system become reliable.
 So on failure, a backup file is available.
 In case of replication all copies should be consistent.
Scalability
 Distributed systems must be scalable as the number of user increases.
 A system is said to be scalable if it can handle the addition of users and resources without
suffering a noticeable loss of performance or increase in administrative complexity.
 Scalability has 3 dimensions:
o Size: Number of users and resources to be processed. Problem associated is
overloading.
o Geography: Distance between users and resources. Problem associated is
communication reliability
o Administration: As the size of distributed systems increases, many of the system needs
to be controlled. Problem associated is administrative mess.
Flexibility
 The design of Distributed operating system should be flexible due to following reasons:
o Ease of Modification: It should be easy to incorporate changes in the system in a user
transparent manner or with minimum interruption caused to the users.
o Ease of Enhancement: New functionality should be added from time to time to make it
more powerful and easy to use.
 A group of users should be able to add or change the services as per the comfortability of their
use.
Performance
 A performance should be better than or at least equal to that of running the same application
on a single-processor system.
 Some design principles considered useful for better performance are as below:
o Batch if possible: Batching often helps in improving performance.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 11

 Unit: 1 Introduction to
Distributed System
o Cache whenever possible: Caching of data at client’s side frequently improves overall
system performance.
o Minimize copying of data: Data copying overhead involves a substantial CPU cost of
many operations.
o Minimize network traffic: It can be improved by reducing internode communication
costs.
Heterogeneity
 This term means the diversity of the distributed systems in terms of hardware, software,
platform, etc.
 Modern distributed systems will likely span different:
o Hardware devices: computers, tablets, mobile phones, embedded devices, etc.
o Operating System: MS Windows, Linux, Mac, Unix, etc.
o Network: Local network, the Internet, wireless network, satellite links, etc.
o Programming languages: Java, C/C++, Python, PHP, etc.
o Different roles of software developers, designers, system managers.
Security
 System must be protected against destruction and unauthorized access.
 Enforcement of Security in a distributed system has the following additional requirements as
compared to centralized system:
o Sender of message should know that message is delivered.
o Receiver of the message should know that the message was sent by genuine sender.
o Both sender and receiver should be guaranteed that the content of message were not
changed while it is in transfer.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 12

 Unit: 2 Communication in
Distributed System
1) What is Computer Network? Give its classification?
 A computer network is a system in which multiple computers are connected to each other to
share information and resources.
 The physical connection between networked computing devices is established using either
cable media or wireless media.
 The best-known computer network is the Internet.

Figure: Computer Network

Local area network (LAN)
 It is privately-owned networks within a single building or campus of up to a few kilometres in
size.
 They are widely used to connect personal computers and workstations in company offices and
factories to share resources (e.g., printers) and exchange information.
 LANs are easy to design and troubleshoot
 In LAN, all the machines are connected to a single cable.
 Different types of topologies such as Bus, Ring, Star and Tree are used.
 The data transfer rates for LAN is up to 10 Gbits/s.
 They transfer data at high speeds. High transmission rate are possible in LAN because of the
short distance between various computer networks.
 They exist in a limited geographical area.

Figure: Local area network (LAN)

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 1

 Unit: 2 Communication in
Distributed System
Wide area Network (WAN)
 WAN spans a large geographical area, often a country or region.
 WAN links different metropolitan’s countries and national boundaries there by enabling easy
communication.
 It may be located entirely with in a state or a country or it may be interconnected around the
world.
 It contains a collection of machines intended for running user (i.e., application) programs. We
will follow traditional usage and call these machines hosts.
 The communication between different users of WAN is established using leased telephone lines
or satellite links and similar channels.

Figure: Wide area network (WAN)

Metropolitan area network (MAN)
 MAN is a larger version of LAN which covers an area that is larger than the covered by LAN but
smaller than the area covered by WAN.
 A metropolitan area network or MAN covers a city.
 The best-known example of a MAN is the cable television network available in many cities.
 MAN connects two or more LANs.

Figure: Metropolitan area network (MAN)

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 2
 Unit: 2 Communication in
Distributed System
Wireless LAN
 WLAN technology allows computer network devices to have network connectivity without the
use of wires, using radio waves.
 These networks use infrared link, Bluetooth or low power radio network operating at a
bandwidth of 1-2 Mbps over a distance of 10m.
 At the top level, wireless networks can be classified as wireless LAN, wireless MAN, and wireless
WAN.
 The wireless LANs are further classified as Personal Area networks (e.g. Bluetooth, wireless
sensor networks) and Business LAN (e.g. 802.11b).

Figure: Wireless LAN

Classification of Network

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 3

 Unit: 2 Communication in
Distributed System
2) Explain OSI Model layered architecture.
OSI Layer Architecture
 OSI model is based on a proposal developed by the International Standards Organization (ISO)
as a first step toward international standardization of the protocols used in the various layers.
 It was revised in 1995.
 The model is called the OSI (Open Systems Interconnection) Reference Model because it deals
with connecting open systems—that is, systems that are open for communication with other
systems.
 The OSI model has seven layers.
o Physical Layer
o Data Link Layer
o Network Layer
o Transport Layer
o Session Layer
o Presentation Layer
o Application Layer

Figure: OSI Reference Model

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 4

 Unit: 2 Communication in
Distributed System
Physical Layer
 The physical layer, the lowest layer of the OSI model, is concerned with the transmission and
reception of the unstructured raw bit stream over a physical medium.
 It describes the electrical/optical, mechanical, and functional interfaces to the physical
medium, and carries the signals for all of the higher layers. It provides:
o Data encoding: modifies the simple digital signal pattern (1s and 0s) used by the PC to
better accommodate the characteristics of the physical medium, and to aid in bit and
frame synchronization.
o Transmission technique: determines whether the encoded bits will be transmitted by
baseband (digital) or broadband (analog) signalling.
o Physical medium transmission: transmits bits as electrical or optical signals appropriate
for the physical medium.
Data link Layer
 The data link layer provides error-free transfer of data frames from one node to another over
the physical layer, allowing layers above it to assume virtually error-free transmission over the
link.
 To do this, the data link layer provides:
o Link establishment and termination: establishes and terminates the logical link between
two nodes.
o Frame traffic control: tells the transmitting node to "back-off" (stop) when no frame
buffers are available.
o Frame sequencing: transmits/receives frames sequentially.
o Frame acknowledgment: provides/expects frame acknowledgments. Detects and
recovers from errors that occur in the physical layer by retransmitting non-
acknowledged frames and handling duplicate frame receipt.
o Frame delimiting: creates and recognizes frame boundaries.
o Frame error checking: checks received frames for integrity.
o Media access management: determines when the node "has the right" to use the
physical medium.
Network Layer
 The network layer controls the operation of the subnet, deciding which physical path the data
should take based on network conditions, priority of service, and other factors.
 To do this, the data link layer provides:
o Routing: routes frames among networks.
o Subnet traffic control: routers (network layer intermediate systems) can instruct a
sending station to "throttle back" its frame transmission when the router's buffer fills
up.
o Frame fragmentation: if it determines that a downstream router's maximum
transmission unit (MTU) size is less than the frame size, a router can fragment a frame
for transmission and re-assembly at the destination station.
o Logical-physical address mapping: translates logical addresses or names, into physical
addresses.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 5

 Unit: 2 Communication in
Distributed System
o Subnet usage accounting: has accounting functions to keep track of frames forwarded
by subnet intermediate systems, to produce billing information.
Transport Layer
 The transport layer ensures that messages are delivered error-free, in sequence, and with no
losses or duplications. It relieves (release) the higher layer protocols from any concern with the
transfer of data between them and their peers.
 The size and complexity of a transport protocol depends on the type of service it can get from
the network layer. For a reliable network layer with virtual circuit capability, a minimal
transport layer is required. If the network layer is unreliable and/or only supports datagrams,
the transport protocol should include extensive error detection and recovery.
 The transport layer provides:
o Message segmentation: accepts a message from the (session) layer above it, splits the
message into smaller units (if not already small enough), and passes the smaller units
down to the network layer. The transport layer at the destination station reassembles
the message.
o Message acknowledgment: provides reliable end-to-end message delivery with
acknowledgments.
o Message traffic control: tells the transmitting station to "back-off" when no message
buffers are available.
 Typically, the transport layer can accept relatively large messages, but there are strict message
size limits imposed by the network (or lower) layer. Consequently, the transport layer must
break up the messages into smaller units, or frames, prepending a header to each frame.
 The transport layer header information must then include control information, such as message
start and message end flags, to enable the transport layer on the other end to recognize
message boundaries.
 In addition, if the lower layers do not maintain sequence, the transport header must contain
sequence information to enable the transport layer on the receiving end to get the pieces back
together in the right order before handing the received message up to the layer above.
Session Layer
 The session layer allows session establishment between processes running on different
stations. It provides:
o Session establishment, maintenance and termination: allows two application processes
on different machines to establish, use and terminate a connection, called a session.
o Session support: performs the functions that allow these processes to communicate
over the network, performing security, name recognition, logging, and so on.
Presentation Layer
 The presentation layer formats the data to be presented to the application layer. It can be
viewed as the translator for the network. This layer may translate data from a format used by
the application layer into a common format at the sending station, then translate the common
format to a format known to the application layer at the receiving station.
 The presentation layer provides:
o Character code translation: for example, ASCII to EBCDIC.
o Data conversion: bit order, CR-CR/LF, integer-floating point, and so on.
Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 6
 Unit: 2 Communication in
Distributed System
o Data compression: reduces the number of bits that need to be transmitted on the
network.
o Data encryption: encrypt data for security purposes. For example, password encryption.
Application Layer
 The application layer serves as the window for users and application processes to access
network services.
 This layer contains a variety of commonly needed functions:
o Resource sharing and device redirection
o Remote file access
o Remote printer access
o Inter-process communication
o Network management
o Directory services
o Electronic messaging (such as mail)
o Network virtual terminals

3) Explain TCP/IP Reference model.

 Transmission Control Protocol/Internet Protocol (TCP/IP) protocol suite is the engine for the
Internet and networks worldwide.
 TCP/IP either combines several OSI layers into a single layer, or does not use certain layers at
all.
 TCP/IP is a set of protocols developed to allow cooperating computers to share resources
across the network.
 The TCP/IP model has five layers.
o Application Layer
o Transport Layer
o Internet Layer
o Data Link Layer
o Physical Network

Figure: TCP/IP Reference Model

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 7
 Unit: 2 Communication in
Distributed System

 As we can see from the above figure, presentation and session layers are not there in TCP/IP
model. Also note that the Network Access Layer in TCP/IP model combines the functions of
Data Link Layer and Physical Layer.
Application Layer
 Application layer is the top most layer of four layer TCP/IP model.
 Application layer is present on the top of the Transport layer.
 Application layer defines TCP/IP application protocols and how host programs interface with
Transport layer services to use the network.
 Application layer includes all the higher-level protocols like DNS (Domain Naming System),
HTTP (Hypertext Transfer Protocol), Telnet, SSH, FTP (File Transfer Protocol), TFTP (Trivial File
Transfer Protocol), SNMP (Simple Network Management Protocol), SMTP (Simple Mail Transfer
Protocol), DHCP (Dynamic Host Configuration Protocol), X Windows, RDP (Remote Desktop
Protocol) etc.
Transport Layer
 The purpose of Transport layer is to permit devices on the source and destination hosts to carry
on a conversation.
 Transport layer defines the level of service and status of the connection used when transporting
data.
 The transport layer provides the end-to-end data transfer by delivering data from an
application to its remote peer.
 The most-used transport layer protocol is the Transmission Control Protocol (TCP), which
provides:
o Reliable delivery data
o Duplicate data suppression
o Congestion control
o Flow control
 Another transport layer protocol is the User Datagram Protocol (UDP), which provides:
o Connectionless
o Unreliable
o Best-effort service
 UDP is used by applications that need a fast transport mechanism and can tolerate the loss of
some data.
Network Layer (Internet Layer)
 The internet layer also called the network layer.
 Internet layer pack data into data packets known as IP datagrams, which contain source and
destination address (logical address or IP address) information that is used to forward the
datagrams between hosts and across networks.
 The Internet layer is also responsible for routing of IP datagrams.
 Internet Protocol (IP) is the most important protocol in this layer.
 It is a connectionless protocol that does not assume reliability from lower layers. IP does not
provide reliability, flow control or error recovery.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 8

 Unit: 2 Communication in
Distributed System
 IP provides a routing function that attempts to deliver transmitted messages to their
destination.
 These message units in an IP network are called an IP datagram.
 Example: IP, ICMP, IGMP, ARP, and RARP.
Network Interface Layer (Network Access Layer)
 Network Access Layer defines details of how data is physically sent through the network,
including how bits are electrically or optically signalled by hardware devices that interface
directly with a network medium, such as coaxial cable, optical fiber, or twisted pair copper wire.
 The protocols included in Network Access Layer are Ethernet, Token Ring, FDDI, X.25, Frame
Relay etc.

OSI(Open System Interconnection) TCP/IP (Transmission Control Protocol/

Internet Protocol)
OSI provides layer functioning and also TCP/IP model is more based on protocols and
defines functions of all the layers. protocols are not flexible with other layers.
In OSI model the transport layer guarantees In TCP/IP model the transport layer does not
the delivery of packets guarantees delivery of packets.
Follows horizontal approach Follows vertical approach.
OSI model has a separate presentation layer TCP/IP doesn’t have a separate presentation
layer
OSI is a general model. TCP/IP model cannot be used in any other
application.
Network layer of OSI model provide both The Network layer in TCP/IP model provides
connection oriented and connectionless connectionless service.
service.
OSI model has a problem of fitting the TCP/IP model does not fit any protocol
protocols in the model
Protocols are hidden in OSI model and are In TCP/IP replacing protocol is not easy.
easily replaced as the technology changes.
OSI model defines services, interfaces and In TCP/IP it is not clearly separated its
protocols very clearly and makes clear services, interfaces and protocols.
distinction between them.
It has 7 layers It has 4 layers

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 9

 Unit: 2 Communication in
Distributed System
4) Explain Inter process communication (IPC) in message passing system.
 Inter process communication (IPC) basically requires information sharing among two or more
processes.

Figure: Inter process communication (a) Shared data approach (b) Message passing approach
 Two basic methods for information sharing are as follows:
o Original sharing, or shared-data approach.
o Copy sharing, or message-passing approach.
 In the shared-data approach, the information to be shared is placed in a common memory area
that is accessible to all processes involved in an IPC.
 In the message-passing approach, the information to be shared is physically copied from the
sender process’s space to the address space of all the receiver processes.
 A message-passing system is a subsystem of distributed operating system that provides a set
of message-based IPC protocols, and does so by shielding the details of complex network
protocols and multiple heterogeneous platforms from programmers.
 It enables processes to communicate by exchanging messages and allows programs to be
written by using simple communication primitives, such as send and receive.
5) Explain desirable features of a good message passing system.
Simplicity
 A message passing system should be simple and easy to use.
 It must be straight forward to construct new applications and to communicate with existing
one by using the primitives provided by message passing system.
Uniform Semantics
 In a distributed system, a message-passing system may be used for the following two types of
inter process communication:
 Local communication, in which the communicating processes are on the same node.
 Remote communication, in which the communicating processes are on different nodes.
 Semantics of remote communication should be as close as possible to those of local
communications.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 10

 Unit: 2 Communication in
Distributed System
Efficiency
 An IPC protocol of a message-passing system can be made efficient by reducing the number of
message exchanges, as far as practicable, during the communication process.
 Some optimizations normally adopted for efficiency include the following:
 Avoiding the costs of establishing and terminating connections between the same pair of
processes for each and every message exchange between them.
 Minimizing the costs of maintaining the connections;
 Piggybacking of acknowledgement of previous messages with the next message during a
connection between a sender and a receiver that involves several message exchanges.
Reliability
 A good IPC protocol can cope with failure problems and guarantees the delivery of a message.
 Handling of lost messages involves acknowledgement and retransmission on the basis of
timeouts.
 A reliable IPC protocol should also be capable of detecting and handling duplicate messages.
 Correctness
 Correctness is a feature related to IPC protocols for group communication.
 Issues related to correctness are as follows:
o Atomicity: It ensures that every message sent to a group of receivers will be delivered
to either all of them or none of them.
o Ordered delivery: It ensures that messages arrive to all receivers in an order acceptable
to the application.
o Survivability: It guarantees that messages will be correctly delivered despite partial
failures of processes, machines, or communication links.
Flexibility
 Not all applications require the same degree of reliability and correctness of the IPC protocol.
 The IPC protocols of a message passing system must be flexible enough to cater to the various
needs of different applications.
 The IPC primitives should be such that user have the flexibility to choose and specify types and
levels of reliability and correctness requirement of their applications.
 IPC primitives must also have the flexibility to permit any kind of control flow between the
cooperating processes, including synchronous and asynchronous send/receive.
Security
 A good message passing system must also be capable of providing a secure end to end
communication.
 Steps necessary for secure communication include the following:
 Authentication of the receiver of a message by the sender.
 Authentication of the sender of a message by its receivers.
 Encryption of a message before sending it over the network.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 11

 Unit: 2 Communication in
Distributed System
Portability
 The message passing system should itself be portable.
 It should be possible to easily construct a new IPC facility on another system by reusing the
basic design of the existing message passing system.
 The application written by using primitives of IPC protocols of the message passing system
should be portable.
 This may require use of an external data representation format for the communication taking
place between two or more processes running on computers of different architectures.
6) What are the Issues in IPC by Message Passing.
 A message is a block of information formatted by a sending process in such a manner that it is
meaningful to the receiving process.
 It consists of a fixed-length header and a variable-size collection of typed data objects.

Figure: A typical message structure.

 The header usually consists of the following elements:
 Address:
o It contains characters that uniquely identify the sending and receiving processes in the
network.
 Sequence number.
o This is the message identifier (ID), which is very useful for identifying lost messages and
duplicate messages in case of system failures.
 Structural information.
o This element also has two parts.
o The type part specifies whether the data to be passed on to the receiver is included
within the message or the message only contains a pointer to the data, which is stored
somewhere outside the contiguous portion of the message.
o The second part of this element specifies the length of the variable-size message data.
 In the design of the IPC protocol for message passing system, the following important issues
need to be considered:
o Who is the sender?
o Who is the receiver?
o Is there one receiver or many receivers?
o Is the message guaranteed to have been accepted by receivers?
o Does the sender need to wait for the reply?
Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 12
 Unit: 2 Communication in
Distributed System
o What should be done if the catastrophic event such as node crash or a communication
link failure occurs during the course of communication?
o What should be done if the receiver is not ready to accept the message: will the
message be discarded or stored in a buffer? In the case of buffering what would be done
if the buffer is full?
o If there are outstanding messages for a receiver, can it choose the order in which to
service the outstanding messages?
7) Explain Buffering in standard message passing model.
 In the standard message passing model, messages can be copied many times: from the user
buffer to the kernel buffer (the output buffer of a channel), from the kernel buffer of the
sending computer (process) to the kernel buffer in the receiving computer (the input buffer of
a channel), and finally from the kernel buffer of the receiving computer (process) to a user
buffer.
 Buffering can be of following types.
Null Buffer (No Buffering)
 In this case, there is no place to temporarily store the message.
 Hence one of the following implementation strategies may be used:
o The message remains in the sender process’s address space and the execution of the
send is delayed until the receiver executes the corresponding receive.
o The message is simply discarded and the time-out mechanism is used to resend the
message after a timeout period.
o The sender may have to try several times before succeeding.

Figure: Message transfer with (a) no buffering (b) single message buffer (c) multi buffer
Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 13
 Unit: 2 Communication in
Distributed System
Single Message Buffer
 In single-message buffer strategy, a buffer having a capacity to store a single message is used
on the receiver’s node.
 This strategy is usually used for synchronous communication; an application module may have
at most one message outstanding at a time.
Finite Bound Buffer or Multi buffer
 Systems using asynchronous mode of communication use finite-bound buffers, also known as
multiple-message buffers.
 In this case message is first copied from the sending process’s memory into the receiving
process’s mailbox and then copied from the mailbox to the receiver’s memory when the
receiver calls for the message.
 When the buffer has finite bounds, a strategy is also needed for handling the problem of a
possible buffer overflow.
 The buffer overflow problem can be dealt with in one of the following two ways:
 Unsuccessful communication.
o In this method, message transfers simply fail, whenever there is no more buffer space
and an error is returned.
 Flow-controlled communication.
o The sender is blocked until the receiver accepts some messages, thus creating space in
the buffer for new messages.
o This method introduces a synchronization between the sender and the receiver and
may result in unexpected deadlocks.
o Due to the synchronization imposed, the asynchronous send does not operate in the
truly asynchronous mode for all send commands.
Unbounded Capacity Buffer
 In the asynchronous mode of communication, since a sender does not wait for the receiver to
be ready, there may be several pending messages that have not yet been accepted by the
receiver.
 Therefore, an unbounded-capacity message-buffer that can store all unreceived messages is
needed to support asynchronous communication with the assurance that all the messages sent
to the receiver will be delivered.
 Unbounded capacity of a buffer is practically impossible.
5) Explain Failure Handling in IPC.
 During inter process communication partial failures such as a node crash or communication link
failure may lead to the following problems:
Loss of request message
 This may happen either due to the failure of communication link between the sender and
receiver or because the receiver’s node is down at the time the request message reaches there.
Loss of response message
 This may happen either due to the failure of communication link between the sender and
receiver or because the sender’s node is down at the time the response message reaches there.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 14

 Unit: 2 Communication in
Distributed System

Figure: Possible problems in IPC when (a) Request message is lost (b) Response message is lost
(c) Receiver’s computer crashed.
Four-message reliable IPC protocol for client-server communication between two processes
 The client sends a request message to the server.
 When the request message is received at the server’s machine, the kernel of that machine
returns an acknowledgment message to the kernel of the client machine.

Figure: Four message reliable IPC protocol for client server

 If the acknowledgment is not received within the timeout period, the kernel of the client
machine retransmits the request message.
 When the server finishes processing the client’s request it returns a reply message to the client.
 When the reply is received at client machine, the kernel of that machine returns an
acknowledgment message to the kernel of the server machine.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 15

 Unit: 2 Communication in
Distributed System
 If the acknowledgment message is not received within the timeout period, the kernel of the
server machine retransmits the reply message.
Three message reliable IPC message protocol
 The client sends a request message to the server.
 When the server finishes processing the client’s request, it returns a reply message (containing
the result of processing) to the client.
 The client remains blocked until the reply is received.

Figure: Three message reliable IPC protocol for client server

 If the reply is not received within the timeout period, the kernel of the client machine
retransmits the request message.
 When the reply message is received at the client’s machine, the kernel of that machine returns
an acknowledgment message to the kernel of the sever machine.
 If the acknowledgment message is not received within the timeout period, the kernel of the
server machine retransmits the reply message.
 If the request message is lost, it will be retransmitted only after the timeout period, which has
been set to a large value to avoid unnecessary retransmission of the request message.
 On the other hand, if the timeout value is not set properly taking into consideration the long
time needed for request processing, unnecessary retransmissions of the request message will
take place.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 16

 Unit: 2 Communication in
Distributed System
Two message reliable IPC message protocol
 The client sends a request message to the server and remains blocked until a reply is received
from the server.

Figure: Two message reliable IPC protocol for client server

 When the server finishes processing the client’s request, it returns a reply message (containing
the result of processing) to the client.
 If the reply is not received within the timeout period, the kernel of the client machine
retransmits the request message.
6) Explain basic concept of client server model.
 The structure of the operation system is like a group of cooperating processes called the
servers, which offer services to the users called the clients.
 Both run on the same microkernel as the user processes.
 A machine can run as single/multiple clients or single/multiple servers, or a mix of both.
 This model depicted in Figure2.23 uses the connectionless Request Reply protocol thus
reducing the overhead of the connection oriented TCP/IP or OSI protocol.

Figure: Client Server model

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 17

 Unit: 2 Communication in
Distributed System
 The Request Reply protocol works as follows:
o The client requesting the service sends a request message to the server.
o The server completes the task and returns the result indicating that it has performed
the task or returns an error message indicating that it has not performed the task.

Figure: The Client server interaction

o Figure shows the client server interaction.
o Here reply serves as acknowledgement to the request.
 As seen in the above figure(b), only three layers of the protocol are used.
 The physical and the data link layer are responsible for transferring packets between the client
and the server through hardware.
 There is no need for routing and establishment of connection.
 The Request Reply protocol defines the set of request and replies to the corresponding
requests. The session management and other higher layers are not required.
7) Write a note on Implementation of Client Server Model.

Figure: Common packet message sequences

 All networks have a maximum packet size.
 If the message is large, it is split into multiple packets and each packet is transmitted
individually.
 What happens if the packets are lost or garbled or arrive at the server out of sequence?
Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 18
 Unit: 2 Communication in
Distributed System
 Solution is assigning a sequence number to each packet depending on the order of the packets.
 In case of multiple packets per message, should the receiver acknowledge each packet or just
the last packet?
 If an acknowledgement is sent for each packet, then only lost packets are retransmitted, but it
increases the message traffic.
 The alternative is to send a single acknowledgement for all the packets.
 This results in transmission of fewer acknowledgements.
 The client may not understand which packet is lost and hence may end up retransmitting all
packets.
 The choice of the method depends on the loss rate of the network.
 Following are different types of packets transmitted across the network
o REQ: Request packet is used to send the request from the client to the server.
o Reply: This message is used to carry the result from the server to the client.
o ACK: Acknowledgement packet is used to send the correct receipt of the packet to the
sender.
o Are You Alive (AYA)? This packet is sent in case the server takes a long time to complete
the client's request.
o I am Alive (IAA): The server, if active, replies with this packet.
 It implies that the client kernel should wait for the reply.
 If no response is received for an AYA message, the client resends the message.
 Figure shows some of the common packet message sequences.
8) Explain Remote procedure call (RPC) and its implementation.

 Remote Procedure Call (RPC) is a protocol that one program can use to request a service from
a program located in another computer on a network without having to understand the
network's details.
 A procedure call is also sometimes known as a function call or a subroutine call.
 RPC uses the client-server model.

Figure: RPC Model

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 19

 Unit: 2 Communication in
Distributed System
 It includes mainly five elements:
o The Client
o The Client stub (stub: Piece of code used for converting parameters)
o The RPCRuntime (RPC Communication Package)
o The Server stub
o The Server
The Client
 It is user process which initiates a remote procedure call.
 The client makes a perfectly normal call that invokes a corresponding procedure in the client
stub.
The Client stub
 On receipt of a request it packs a requirement into a message and asks to RPCRuntime to send.
 On receipt of a result it unpacks the result and passes it to client.
RPCRuntime
 It handles transmission of messages between client and server.
The Server stub
 It unpacks a call request and make a perfectly normal call to invoke the appropriate procedure
in the server.
 On receipt of a result of procedure execution it packs the result and asks to RPCRuntime to
send.
The Server
 It executes an appropriate procedure and returns the result from a server stub.

Figure: Implementation of RPC mechanism

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 20
 Unit: 2 Communication in
Distributed System
Steps of Implementation are as follows:
1) The client calls a local procedure, called the client stub.
2) Network messages are sent by the client stub to the remote system (via a system call to the
local kernel using sockets interfaces).
3) Network messages are transferred by the kernel to the remote system via some protocol
(either connectionless or connection-oriented).
4) A server stub, sometimes called the skeleton, receives the messages on the server.
It unmarshals the arguments from the messages.
5) The server stub calls the server function, passing it the arguments that it received from the
client.
6) When server function is finished, it returns to the server stub with its return values.
7) The server stub converts the return values, if necessary, and marshals them into one or
more network messages to send to the client stub.
8) Messages get sent back across the network to the client stub.
9) The client stub reads the messages from the local kernel.
10) The client stub then returns the results to the client function, converting them from the
network representation to a local one if necessary.

9) Explain Implementation Issues in Remote procedure call (RPC).

RPC protocols
 The first issue is the choice of the RPC protocol.
Connection oriented protocol
o The client is bound to the server and a connection is established between them.
o The advantage: communication becomes much easier.
o The disadvantage: especially over a LAN, is the performance loss.
Connectionless protocol
o IP and UDP are easy to use and fit in well with existing UNIX systems and networks such
as the Internet, but downside is performance.
 Packet and message length is another issue.
 Doing an RPC has a large, fixed overhead, independent of the amount of data sent.
 Thus reading a 64K file in a single 64K RPC is vastly more efficient than reading it in 64 1K RPCs.
 It is therefore important that the protocol and network should allow large transmissions.
 Some RPC systems are limited to small sizes.
 In addition, many networks cannot handle large packets so a single RPC will have to be split
over multiple packets, causing extra overhead.
Acknowledgement
 If RPCs are broken up into many small packets, a new issue arises: Should individual packets be
acknowledged or not?

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 21

 Unit: 2 Communication in
Distributed System
 Suppose, for example, that a client wants to write a 4K block of data to a file server, but the
system cannot handle packets larger than 1K.

Figure: Acknowledgement to RPC messages

 One strategy, known as a stop-and-wait protocol, is for the client to send packet 0 with the first
1K, then wait for an acknowledgement from the server, as illustrated in Figure (b).
 Then the client sends the second 1K, waits for another acknowledgement, and so on.
 The alternative, often called a blast protocol, is simply for the client to send all the packets as
fast as it can.
 With this method, the server acknowledges the entire message when all the packets have been
received, not one by one.
 The blast protocol is illustrated in Figure (c).
Critical path
 The sequence of instructions that is executed on every RPC is called the critical path, and is
shown in Figure.

Figure: Critical path from client to server

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 22
 Unit: 2 Communication in
Distributed System
 Even for the null RPC, the dominant costs are the context switch to the server stub when a
packet arrives, the interrupt service routine, and moving the packet to the network interface
for transmission.
 For the 1440-byte RPC, the picture changes considerably, with the Ethernet transmission time
now being the largest single component, with the time for moving the packet into and out of
the interface coming in close behind.
Copying
 An issue that frequently dominates RPC execution times is copying.
 The number of times a message must be copied varies from 1 to about 8, depending on the
hardware, software and type of call.
 In the best case, the network chip can DMA the message directly out of the client stub's address
space onto the network (copy 1), depositing it in the server kernel's memory in real.
 In the worst case, the kernel copies the message from the client stub into a kernel buffer for
subsequent transmission, either because it is not convenient to transmit directly from user
space or the network is currently busy (copy 1).
 Later, the kernel copies the message, in software, to a hardware buffer on the network
interface board (copy 2).
 At this point, the hardware is started, causing the packet to be moved over the network to the
interface board on the destination machine (copy 3).
 When the packet-arrived interrupt occurs on the server's machine, the kernel copies it to a
kernel buffer (copy 4).
 Finally, the message has to be copied to the server stub (copy 5).
 Suppose that it takes an average of 500 nsec to copy a 32-bit word; then with eight copies, each
word needs 4 msec, no matter how fasts the network itself is.
Timer management
 Huge amount of machine time is wasted in managing the timers.
 Setting a timer requires building a data structure specifying when the timer is to expire and
what is to be done after timer expires.
 The data structure is then inserted into a list consisting of the other pending timers.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 23

 Unit: 3 Synchronization
1) What is Synchronization? Explain Clock Synchronization and implementation of computer clock.
 Synchronization is coordination with respect to time, and refers to the ordering of events and
execution of instructions in time.
 It is often important to know when events occurred and in what order they occurred.
 Clock synchronization deals with understanding the temporal ordering of events produced by
concurrent processes.
 It is useful for synchronizing senders and receivers of messages, controlling joint activity, and
the serializing concurrent access to shared objects.
 The goal is that multiple unrelated processes running on different machines should be in
agreement with and be able to make consistent decisions about the ordering of events in a
system.
 Another aspect of clock synchronization deals with synchronizing time-of-day clocks among
groups of machines.
 In this case, we want to ensure that all machines can report the same time, regardless of how
imprecise their clocks may be or what the network latencies are between the machines.
 A computer clock usually consists of three components-a quartz crystal that oscillates at a well-
defined frequency, a counter register, and a constant register.
 The constant register is used to store a constant value that is decided based on the frequency
of oscillation of the quartz crystal.
 The counter register is used to keep track of the oscillations of the quartz crystal.
 That is, the value in the counter register is decremented by 1 for each oscillation of the quartz
crystal.
 When the value of the counter register becomes zero, an interrupt is generated and its value
is reinitialized to the value in the constant register.
 Each interrupt is called a clock tick.
 To make the computer clock function as an ordinary clock used by us in our day-today life, the
following things are done:
o The value in the constant register is chosen so that 60 clock ticks occur in a second.
o The computer clock is synchronized with real time (external clock). For this, two more
values are stored in the system-a fixed starting date and time and the number of ticks.
o For example, in UNIX, time begins at 0000 on January 1, 1970.
o At the time of initial booting, the system asks the operator to enter the current date
and time.
o The system converts the entered value to the number of ticks after the fixed starting
date and time.
o At every clock tick, the interrupt service routine increments the value of the number of
ticks to keep the clock running.
2) Explain Drifting of Clock.
 A clock always runs at a constant rate because its quartz crystal oscillates at a well-defined
frequency.
 However, due to differences in the crystals, the rates at which two clocks run are normally
different from each other.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 1

 Unit: 3 Synchronization
 The difference in the oscillation period between two clocks might be extremely small, but the
difference accumulated over many oscillations leads to an observable difference in the times
of the two clocks, no matter how accurately they were initialized to the same value.
 Therefore, with the passage of time, a computer clock drifts from the real-time clock that was
used for its initial setting.
 For clocks based on a quartz crystal, the drift rate is approximately 10-6, giving a difference of
1 second every 1,000,000 seconds, or 11.6 days.
 Hence a computer clock must be periodically resynchronized with the real-time clock to keep
it non faulty.
 Even non faulty clocks do not always maintain perfect time.
 A clock is considered non faulty if there is a bound on the amount of drift from real time for
any given finite time interval.
 As shown in Figure, after synchronization with a perfect clock, slow and fast clocks drift in
opposite directions from the perfect clock.
 This is because for slow clocks dC/dt < 1 and for fast clocks dC/dt > 1.

Figure: Slow, perfect, and fast clocks.

 A distributed system requires the following types of clock synchronization:
o Synchronization of the computer clocks with real-time (or external) clocks.
o This type of synchronization is mainly required for real-time applications.
o That is, external clock synchronization allows the system to exchange information about
the timing of events with other systems and users.
o An external time source that is often used as a reference for synchronizing computer
clocks with real time is the Coordinated Universal Time (UTC).
o Mutual (or internal) synchronization of the clocks of different nodes of the system.
o This type of synchronization is mainly required for those applications that require a
consistent view of time across all nodes of a distributed system as well as for the
measurement of the duration of distributed activities that terminate on a node different
from the one on which they start.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 2

 Unit: 3 Synchronization
3) Classify and explain clock synchronization algorithms.

Figure: Classification of Synchronization Algorithms

Centralized clock synchronization algorithms
 In centralized clock synchronization algorithms one node has a real-time receiver.
 This node, called the time server node whose clock time is regarded as correct and used as the
reference time.
 The goal of these algorithms is to keep the clocks of all other nodes synchronized with the clock
time of the time server node.
 Depending on the role of the time server node, centralized clock synchronization algorithms
are again of two types –
o Passive Time Sever algorithm
o Active Time Server algorithm
Distributed clock synchronization algorithms
 Externally synchronized clocks are also internally synchronized
 Each node’s clock is independently synchronized with real time.
 All clocks of system remain mutually synchronized.
 Each node is equipped with real time receiver so that each node can be independently
synchronized.
 Theoretically internal synchronization of clock is not required.
 In practice, due to inherent inaccuracy of real time clocks, different real time clocks produce
different time.
 Internal synchronization is performed for better accuracy.
 Types of internal Synchronization:
o Global averaging
o Localized Averaging

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 3

 Unit: 3 Synchronization
4) Explain Passive time server algorithms.
Passive time server Algorithm
 Each node periodically sends a message called ‘time=?’ to the time server.
 When the time server receives the message, it responds with ‘time=T’ message.
 Assume that client node has a clock time of T0 when it sends ‘time=?’ and time T1 when it
receives the ‘time=T’ message.
 T0 and T1 are measured using the same clock, thus the time needed in propagation of message
from time server to client node would be (T1- T0)/2
 When client node receives the reply from the time server, client node is readjusted to
Tserver+(T1-T0)/2.
 Two methods have been proposed to improve estimated value
o Let the approximate time taken by the time server to handle the interrupt and process
the message request message ‘time=?’ is equal to I.
o Hence, a better estimate for time taken for prapogation of response message from time
server node to client node is taken as (T1- T0-I)/2
o Clock is adjusted to the value Tserver+(T1- T0-I)/2

Figure: Time approximation using passive time server algorithm

 Christian method.
o This method assumes that a certain machine, the time server is synchronized to the UTC
in some fashion called T.
o Periodically all clock is synchronized with time server.
o Other machines send a message to the time server, which responds with T in a
response, as fast as possible.
o The interval (T1- T0) is measured many times.
o Those measurements in which (T1- T0) exceeds a specific threshold values are
considered to be unreliable and are discarded.
o Only those values that fall in the range (T1- T0-2Tmin) are considered for calculating the
correct time.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 4

 Unit: 3 Synchronization
o For all the remaining measurements, an average is calculated which is added to T.
o Alternatively, measurement for which value of (T1- T0) is minimum, is considered most
accurate and half of its value is added to T.

5) Explain Active time server algorithms.

 It is also called Berkeley Algorithm.
 An algorithm for internal synchronization of a group of computers.
 A master polls to collect clock values from the others (slaves).
 The master uses round trip times to estimate the slaves’ clock values.
 It obtains average from participating computers.
 It sends the required adjustment to the slaves.
 If master fails, can elect a new master to take over.
 It Synchronizes all clocks to average.

Figure: (a) The time daemon asks all the other machines for their clock values. (b) The machines
answer. (c) The time daemon tells everyone how to adjust their clock.
 Here the time server (actually, a time daemon) is active, polling every machine periodically to
ask what time it is there.
 Based on the answers, it computes an average time and tells all the other machines to advance
their clocks to the new time or slow their clocks down until some specified reduction has been
achieved.
 In Figure (a), at 3:00, the time daemon tells the other machines its time and asks for theirs.
 In Figure (b), they respond with how far ahead or behind the time daemon they are.
 Armed with these numbers, the time daemon computes the average and tells each machine
how to adjust its clock Figure (c).

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 5

 Unit: 3 Synchronization
6) Explain Global averaging algorithm and Local averaging algorithm.
Global averaging algorithm
 One class of decentralized clock synchronization algorithms works by dividing time into fixed-
length resynchronization intervals.
 The ith interval starts at T0+iR and runs until T0+(i+1) R, where T0 is an agreed upon moment
in the past, and R is a system parameter.
 At the beginning of each interval, every machine broadcasts the current time according to its
clock.
 Because the clocks on different machines do not run at exactly the same speed, these
broadcasts will not happen precisely simultaneously.
 After a machine broadcasts its time, it starts a local timer to collect all other broadcasts that
arrive during some interval 5.
 When all the broadcasts arrive, an algorithm is run to compute a new time from them.
 The simplest algorithm is just to average the values from all the other machines.
 A slight variation on this theme is first to discard the m highest and m lowest values, and
average the rest.
 Discarding the extreme values can be regarded as self-defense against up to m faulty clocks
sending out nonsense.
 Another variation is to try to correct each message by adding to it an estimate of the
propagation time from the source.
 This estimate can be made from the known topology of the network, or by timing how long it
takes for probe messages to be enclosed.
Local averaging algorithm
 This algorithm attempt to overcome the drawbacks of the global averaging algorithm.
 Nodes of a distributed system are logically arranged in some kind of pattern such as Ring.
 Periodically each node exchanges its clock time with his neighbors in the ring.
 Then its sets clock time by taking Average.
 Average is calculated by, its own clock time and clock times of its neighbors.

7) What is Mutual Exclusion?

 Mutual Exclusion is a process that prevents multiple threads or processes from accessing
shared resources at the same time.
 Concurrent access of processes to a shared resource or data is executed in mutually exclusive
manner.
 Only one process is allowed to execute the critical section (CS) at any given time.
 A critical section is a section in a program that accesses shared resources.
 In a distributed system, shared variables or a local kernel cannot be used to implement mutual
exclusion.
 Message passing is the sole means for implementing distributed mutual exclusion.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 6

 Unit: 3 Synchronization
8) Explain Centralized Mutual Exclusion algorithm.

Figure: (a) Process 1 asks the coordinator for permission to enter a critical
region. Permission is granted. (b) Process 2 then asks permission to enter the
same critical region. The coordinator does not reply. (c) When process 1 exits
the critical region, it tells the coordinator, which then replies to 2.

 The most straight forward way to achieve mutual exclusion in a distributed system is to
simulate how it is done in a one-processor system.
 One process is elected as the coordinator (e.g., the one running on the machine with the
highest network address).
 Whenever a process wants to enter a critical region, it sends a request message to the
coordinator stating which critical region it wants to enter and asking for permission.
 If no other process is currently in that critical region, the coordinator sends back a reply
granting permission.
 When the reply arrives, the requesting process enters the critical region.
 Now suppose that another process 2 in figure asks for permission to enter the same critical
region.
 The coordinator knows that a different process is already in the critical region, so it cannot
grant permission.
 The exact method used to deny permission is system dependent.
 In Fig.(b), the coordinator just refrains from replying, thus blocking process 2, which is waiting
for a reply.
 Alternatively, it could send a reply saying ‘‘permission denied.’’ Either way, it queues the
request from 2 for the time being and waits for more messages.
 When process 1 exits the critical region, it sends a message to the coordinator releasing its
exclusive access, as shown in Fig.(c).
 The coordinator takes the first item off the queue of deferred requests and sends that process
a grant message.
 If the process was still blocked (i.e., this is the first message to it), it unblocks and enters the
critical region.
 If an explicit message has already been sent denying permission, the process will have to poll
for incoming traffic or block later.
 Either way, when it sees the grant, it can enter the critical region.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 7

 Unit: 3 Synchronization
9) Explain Distributed Mutual Exclusion algorithm.
 When a process wants to enter a critical region, it builds a message containing the name of the
critical region it wants to enter, its process number, and the current time.
 It then sends the message to all other processes, conceptually including itself.
 The sending of messages is assumed to be reliable; that is, every message is acknowledged.
 Reliable group communication if available, can be used instead of individual messages.
 When a process receives a request message from another process, the action it takes depends
on its state with respect to the critical region named in the message.
 Three cases have to be distinguished:
o If the receiver is not in the critical region and does not want to enter it, it sends back an
OK message to the sender.
o If the receiver is already in the critical region, it does not reply. Instead, it queues the
request.
o If the receiver wants to enter the critical region but has not yet done so, it compares
the timestamp in the incoming message with the one contained in the message that it
has sent everyone.
o The lowest one wins. If the incoming message is lower, the receiver sends back an OK
message.
o If its own message has a lower timestamp, the receiver queues the incoming request
and sends nothing.

Figure (a) Two processes want to enter the same critical region at the same
moment. (b) Process 0 has the lowest timestamp, so it wins. (c) When process 0
is done, it sends an OK also, so 2 can now enter the critical region.
 Process 0 sends everyone a request with timestamp 8.
 while at the same time, process 2 sends everyone a request with timestamp 12.
 Process 1 is not interested in entering the critical region, so it sends OK to both senders.
 Processes 0 and 2 both see the conflict and compare timestamps.
 Process 2 sees that it has lost, so it grants permission to 0 by sending OK.
 Process 0 now queues the request from 2 for later processing and enters the critical region.
 When it is finished, it removes the request from 2 from its queue and sends an OK message to
process 2, allowing the latter to enter its critical region.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 8

 Unit: 3 Synchronization
10) Explain Token Ring Algorithm in Mutual Exclusion.
 Here we have a bus network, as shown in Fig. (a), (e.g., Ethernet), with no inherent ordering
of the processes.
 In software, a logical ring is constructed in which each process is assigned a position in the
ring, as shown in Fig.(b).
 The ring positions may be allocated in numerical order of network addresses or some other
means.
 It does not matter what the ordering is. All that matters is that each process knows who is
next in line after itself.

Figure (a) An unordered group of processes on a network. (b) A logical

ring constructed in software.
 When the ring is initialized, process 0 is given a token.
 The token circulates around the ring.
 It is passed from process k to process k +1 (modulo the ring size) in point-to-point messages.
 When a process acquires the token from its neighbor, it checks to see if it is attempting to enter
a critical region.
 If so, the process enters the region, does all the work it needs to, and leaves the region.
 After it has exited, it passes the token along the ring. It is not permitted to enter a second
critical region using the same token.
 If a process is handed the token by its neighbor and is not interested in entering a critical region,
it just passes it along.
 As a consequence, when no processes want to enter any critical regions, the token just
circulates at high speed around the ring.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 9

 Unit: 3 Synchronization
11) What is Deadlock?
Deadlock
 A set of processes is deadlocked if each process in the set is waiting for an event that only
another process in the set can cause.
 Suppose process D holds resource T and process C holds resource U.
 Now process D requests resource U and process C requests resource T but none of process will
get this resource because these resources are already hold by other process so both can be
blocked, with neither one be able to proceed, this situation is called deadlock.
 Example:

Figure 4-1 Resource allocation graphs. (a) Holding a resource. (b) Requesting a resource. (c)
Deadlock.
 As shown in figure 4-1, resource T assigned to process D and resource U is assigned to process
C.
 Process D is requesting / waiting for resource U and process C is requesting / waiting for
resource T.
 Processes C and D are in deadlock over resources T and U.

12) List the conditions that lead to deadlock.

Conditions that lead to deadlock
 There are four conditions that must hold for deadlock:

1) Mutual exclusion condition

 Each resource is either currently assigned to exactly one process or is available.
2) Hold and wait condition
 Process currently holding resources granted earlier can request more resources.
3) No preemption condition
 Previously granted resources cannot be forcibly taken away from process.
4) Circular wait condition
 There must be a circular chain of 2 or more processes. Each process is waiting for
resource that is held by next member of the chain.
 All four of these conditions must be present for a deadlock to occur.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 10

 Unit: 3 Synchronization
13) Explain deadlock detection and recovery.
Deadlock detection with single resource of each type.
 Algorithm for detecting deadlock for single resource
1. For each node, N in the graph, perform the following five steps with N as the starting node.
2. Initialize L to the empty list, designate all arcs as unmarked.
3. Add current node to end of L, check to see if node now appears in L two times. If it does,
graph contains a cycle (listed in L), algorithm terminates.
4. From given node, see if any unmarked outgoing arcs. If so, go to step 5; if not, go to step 6.
5. Pick an unmarked outgoing arc at random and mark it. Then follow it to the new current
node and go to step 3.
6. If this is initial node, graph does not contain any cycles, algorithm terminates. Otherwise,
dead end. Remove it, go back to previous node, make that one current node, go to step 3.

Figure 4-2 (a) A resource graph. (b) A cycle extracted from (a)
 For example as shown in figure 4-2 (a),
 We are starting from node D.
 Empty list L = ()
 Add current node so Empty list = (D).
 From this node there is one outgoing arc to T so add T to empty list.
 So Empty list become L = (D, T).
 Continue this step….so we get empty list as below
 L = (D, T, E)………… L = (D, T, E, V, G, U, D)
 In the above step in empty list the node D appears twice, so deadlock.
Deadlock detection for multiple resource
 When multiple copies of some of the resources exist, a matrix-based algorithm can be used for
detecting deadlock among n processes.
 Let the number of resource classes be m, with E1 resources of class 1, E2 resources of class 2,
and generally, Ei resources of class i (1 < i < m).
 E is the existing resource vector. It gives the total number of instances of each resource in
existence.
 Let A be the available resource vector, with Ai giving the number of instances of resource i that
are currently available (i.e., unassigned).
Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 11
 Unit: 3 Synchronization
 There are two matrices used in the algorithm. C is a current allocation matrix, and R, the request
matrix.

Figure 4-3. The four data structures needed by the deadlock detection algorithm.
 The i-th row of C tells how many instances of each resource class Pi currently holds.
 Thus Cij is the number of instances of resource j that are held by process i.
 Similarly, Rij is the number of instances of resource j that Pi wants.
 These four data structures are shown in Fig. 4-3.
 The deadlock detection algorithm can now be given, as follows.
1. Look for an unmarked process, Pi, for which the i-th row of R is less than or equal to A.
2. If such a process is found, add the i-th row of C to A, mark the process, and go back to
step 1.
3. If no such process exists, the algorithm terminates.
When the algorithm finishes, all the unmarked processes, if any, are deadlocked.
Example of deadlock detection in multiple resource

Figure 4-4. An example for the deadlock detection algorithm.

 E is the total no of each resource, C is the number of resources held by each process, A is the
number of resources that are available (free), and R is the number of resources still needed by
each process to proceed.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 12

 Unit: 3 Synchronization
 At this moment, we run the algorithm:
 P3 can be satisfied, so P3 completes, returns resources A = (2 2 2 0)
 P2 can be satisfied, so P2 completes, returns resources A = (4 2 2 1)
 P1 can be satisfied, so P1 completes.
 Now all processes are marked, so no deadlock.

Deadlock recovery
 Recovery through preemption
o In some cases, it may be possible to temporarily take a resource away from its
current owner and give it to another process.
o The ability to take a resource away from a process, have another process use it, and
then give it back without the process noticing it is highly dependent on the nature
of the resource.
o Recovering this way is frequently difficult or impossible.
o Choosing the process to suspend depends largely on which ones have resources that
can easily be taken back.
 Recovery through rollback
o Create a checkpoint.
o Checkpoint a process periodically.
o Checkpointing a process means that its state is written to a file so that it can be
restarted later.
o The checkpoint contains not only the memory image, but also the resource state,
that is, which resources are currently assigned to the process.
o When a deadlock is detected, it is easy to see which resources are needed.
o To do the recovery, a process that owns a needed resource is rolled back to a point
in time before it acquired some other resource by starting one of its earlier
checkpoints.
o In effect, the process is reset to an earlier moment when it did not have the
resource, which is now assigned to one of the deadlocked processes.
o If the restarted process tries to acquire the resource again, it will have to wait until
it becomes available.
 Recovery through killing processes
o The crudest, but simplest way to break a deadlock is to kill one or more processes.
o One possibility is to kill a process in the cycle. With a little luck, the other processes
will be able to continue.
o If this does not help, it can be repeated until the cycle is broken.
o Alternatively, a process not in the cycle can be chosen as the victim in order to
release its resources.
o In this approach, the process to be killed is carefully chosen because it is holding
resources that some process in the cycle needs.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 13

 Unit: 3 Synchronization
14) How deadlock can be prevented? OR Explain deadlock prevention.
Deadlock Prevention
 Deadlock can be prevented by attacking the one of the four conditions that leads to deadlock.
1) Attacking the Mutual Exclusion Condition
o No deadlock if no resource is ever assigned exclusively to a single process.
o Some devices can be spooled such as printer, by spooling printer output; several
processes can generate output at the same time.
o Only the printer daemon process uses physical printer.
o Thus deadlock for printer can be eliminated.
o Not all devices can be spooled.
o Principle:
 Avoid assigning a resource when that is not absolutely necessary.
 Try to make sure that as few processes as possible actually claim the
resource.
2) Attacking the Hold and Wait Condition
o Require processes to request all their resources before starting execution.
o A process is allowed to run if all resources it needed is available. Otherwise nothing
will be allocated and it will just wait.
o Problem with this strategy is that a process may not know required resources at
start of run.
o Resource will not be used optimally.
o It also ties up resources other processes could be using.
o Variation: A process must give up all resources before making a new request.
Process is then granted all prior resources as well as the new ones only if all required
resources are available.
o Problem: what if someone grabs the resources in the meantime how can the
processes save its state?
3) Attacking the No Preemption Condition
o This is not a possible option.
o When a process P0 request some resource R which is held by another process P1
then resource R is forcibly taken away from the process P1 and allocated to P0.
o Consider a process holds the printer, halfway through its job; taking the printer
away from this process without having any ill effect is not possible.
4) Attacking the Circular Wait Condition
o To provide a global numbering of all the resources.
o Now the rule is this: processes can request resources whenever they want to, but all
requests must be made in numerical order.
o A process need not acquire them all at once.
o Circular wait is prevented if a process holding resource n cannot wait for resource
m, if m > n.
o No way to complete a cycle.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 14

 Unit: 3 Synchronization

o A process may request 1st a CD ROM drive, then tape drive. But it may not request
1st a plotter, then a Tape drive.
o Resource graph can never have cycle.
15) What is Election Algorithms.
 Several distributed algorithms require a coordinator process in the entire System.
 It performs coordination activity needed for the smooth running of processes.
 Two examples of such coordinator processes encountered here are:
o The coordinator in the centralized algorithm for mutual exclusion.
o The central coordinator in the centralized deadlock detection algorithm.
 If the coordinator process fails, a new coordinator process must be elected to take up the job
of the failed coordinator.
 Election algorithms are meant for electing a coordinator process from the currently running
processes.
 Election algorithms are based on the following assumptions:
o Each process in the system has a unique priority number.
o Whenever an election is held, the process having the highest priority number among
the currently active processes is elected as the coordinator.
o On recovery, a failed process can take appropriate actions to rejoin the set of active
processes.

16) Explain Bully Algorithm for Election.

 Bully algorithm specifies the process with the highest identifier will be the coordinator of the
group. It works as follows:
 When a process p detects that the coordinator is not responding to requests, it initiates an
election:
o p sends an election message to all processes with higher numbers.
o If nobody responds, then p wins and takes over.
o If one of the processes answers, then p’s job is done.
 If a process receives an election message from a lower-numbered process at any time, it:
o sends an OK message back.
o holds an election (unless its already holding one).
 A process announces its victory by sending all processes a message telling them that it is the
new coordinator.
 If a process that has been down recovers, it holds an election.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 15

 Unit: 3 Synchronization

Figure: Bully Algorithm. (a) Process 4 holds an election. (b) Processes 5 and 6 respond, telling 4
to stop. (c) Now 5 and 6 each hold an election. (d) Process 6 tells 5 to stop. (e) Process 6 wins
and tells everyone.
 The group consists of eight processes, numbered from 0 to 7.
 Previously process 7 was the coordinator, but it has just crashed.
 Process 4 is the first one to notice this, so it sends ELECTION messages to all the processes
higher than it, namely 5, 6, and 7, as shown in Figure (a).
 Processes 5 and 6 both respond with OK, as shown in Figure (b).
 Upon getting the first of these responses, 4 knows that its job is over.
 It knows that one of these bigwigs will take over and become coordinator.
 It just sits back and waits to see who the winner will be.
 In Figure (c), both 5 and 6 hold elections, each one only sending messages to those processes
higher than itself.
 In Figure (d) process 6 tells 5 that it will take over.
 At this point 6 knows that 7 is dead and that it (6) is the winner.
 If there is state information to be collected from disk or elsewhere to pick up where the old
coordinator left off, 6 must now do what is needed.
 When it is ready to take over, 6 announces this by sending a COORDINATOR message to all
running processes.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 16

 Unit: 3 Synchronization
 When 4 gets this message, it can now continue with the operation it was trying to do when it
discovered that 7 was dead, but using 6 as the coordinator this time.
 In this way the failure of 7 is handled and the work can continue.
 If process 7 is ever restarted, it will just send all the others a COORDINATOR message and bully
them into submission.
17) Explain Ring Election Algorithm.

 The ring algorithm assumes that the processes are arranged in a logical ring and each process
is knowing the order of the ring of processes.
 If any process detects failure, it constructs an election message with its process ID (e.g., its
network address and local process ID) and sends it to its neighbor.
 If the neighbor is down, the process skips over it and sends the message to the next process in
the ring.
 This process is repeated until a running process is located.
 At each step, the process adds its own process ID to the list in the message and sends the
message to its living neighbor.
 Eventually, the election message comes back to the process that started it.
 The process then picks either the highest or lowest process ID in the list and sends out a
message to the group informing them of the new coordinator.

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 17

 Unit: 3 Synchronization
 Fig.1 shows a ring of six processes. P2 detects that the coordinator P0 is dead.
 It starts an election by sending an election message containing its process ID to its neighbor P3.
(Fig.1)
 P3 receives an election message and sends an election message to its neighbor with the ID of
P3 suffixed to the list. (Fig.2)
 The same sequence of events occurs with P4, which adds its ID to the list it received from P3
and sends an election message to P5.
 P5 then tries to send an election message to P0. (Fig. 3)
 P0 is dead and the message is not delivered, P5 tries again to its neighbor P1. (Fig. 4)
 The message is received by P2, the originator of the election. (Fig. 5)
 P2 recognizes that it is the initiator of the election because its ID is the first in the list of
processes.
 It then picks a leader. it chooses the lowest-numbered process ID, which is that of P1.
 It then informs the rest of the group of the new coordinator (Fig. 6).

Prof. Dipak Ramoliya | 2160710 – Distributed Operating System 18

 Unit: 4 Processes and Processors
in Distributed Systems
1) What is thread? Explain thread structure. OR Explain thread in brief with its structure.
Thread
 A program has one or more locus of execution. Each execution is called a thread of execution.
 In traditional operating systems, each process has an address space and a single thread of
execution.
 It is the smallest unit of processing that can be scheduled by an operating system.
 A thread is a single sequence stream within in a process. Because threads have some of the
properties of processes, they are sometimes called lightweight processes. In a process, threads
allow multiple executions of streams.
Thread Structure
 Process is used to group resources together and threads are the entities scheduled for
execution on the CPU.
 The thread has a program counter that keeps track of which instruction to execute next.
 It has registers, which holds its current working variables.
 It has a stack, which contains the execution history, with one frame for each procedure called
but not yet returned from.
 Although a thread must execute in some process, the thread and its process are different
concepts and can be treated separately.
 What threads add to the process model is to allow multiple executions to take place in the
same process environment, to a large degree independent of one another.
 Having multiple threads running in parallel in one process is similar to having multiple processes
running in parallel in one computer.

Figure: (a) Three processes each with one thread. (b) One process with three threads.
 In former case, the threads share an address space, open files, and other resources.
 In the latter case, process share physical memory, disks, printers and other resources.
 In Fig.(a) we see three traditional processes. Each process has its own address space and a
single thread of control.
 In contrast, in Fig.(b) we see a single process with three threads of control.
 Although in both cases we have three threads, in Fig.(a) each of them operates in a different
address space, whereas in Fig.(b) all three of them share the same address space.
 Like a traditional process (i.e., a process with only one thread), a thread can be in any one of
several states: running, blocked, ready, or terminated.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 1

 Unit: 4 Processes and Processors
in Distributed Systems
 When multithreading is present, processes normally start with a single thread present. This
thread has the ability to create new threads by calling a library procedure thread_create.
 When a thread has finished its work, it can exit by calling a library procedure thread_exit.
 One thread can wait for a (specific) thread to exit by calling a procedure thread_join. This
procedure blocks the calling thread until a (specific) thread has exited.
 Another common thread call is thread_yield, which allows a thread to voluntarily give up the
CPU to let another thread run.

Figure: Each thread has its own stack.

2) What are the Similarities and Dissimilarities between Process and Thread?

Similarities between Process and Thread

 Like processes threads share CPU and only one thread is running at a time.
 Like processes threads within a process execute sequentially.
 Like processes thread can create children.
 Like a traditional process, a thread can be in any one of several states: running, blocked, ready
or terminated.
 Like process threads have Program Counter, Stack, Registers and State.

Dissimilarities between Process and Thread

 Unlike processes threads are not independent of one another.
 Threads within the same process share an address space.
 Unlike processes all threads can access every address in the task.
 Unlike processes threads are design to assist one other. Note that processes might or might
not assist one another because processes may be originated from different users.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 2

 Unit: 4 Processes and Processors
in Distributed Systems
3) Give difference between User level threads and Kernel level threads.

USER LEVEL THREAD KERNEL LEVEL THREAD

User thread are implemented by users. Kernel threads are implemented by OS.
OS doesn’t recognized user level threads. Kernel threads are recognized by OS.
Implementation of User threads is easy. Implementation of Kernel thread is complex.
Context switch time is less. Context switch time is more.
Context switch requires no hardware support. Context switch requires hardware support.
If one user level thread performs blocking If one kernel thread performs blocking
operation, then entire process will be blocked. operation, then another thread with in same
process can continue execution.
Example: Java thread, POSIX threads. Example: Window Solaris

4) Explain Thread models. OR Explain type of Thread Usage.

Dispatcher Worker model

 In this model, the process consists of a single dispatcher thread and multiple worker threads.
 The dispatcher thread accepts requests from clients and, after examining the request,
dispatches the request to one of the free worker threads for further processing of the request.
 Each worker thread works on a different client request.
 Therefore, multiple client requests can be processed in parallel.

Figure: Dispatcher worker model

Team model
 In this model, all threads behave as equal.
 Each threads gets and processes clients requests on its own.
 This model is often used for implementing specialized threads within a process.
 Each thread of the process is specialized in servicing a specific type of request.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 3

 Unit: 4 Processes and Processors
in Distributed Systems

Figure: Team model

Pipeline model
 This model is useful for applications based on the producer-consumer model.
 The output data generated by one part of the application is used as input for another part of
the application.
 In this model, the threads of a process are organized as a pipeline so that the output data
generated by the first thread is used for processing by the second thread, the output of the
second thread is used for processing by the third thread, and so on.
 The output of the last thread in the pipeline is the final output of the process to which the
threads belong.

Figure: Pipeline model

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 4

 Unit: 4 Processes and Processors
in Distributed Systems
5) Explain Designing issues in Thread package.
 A system that supports thread facility must provide a set of primitives to its users for threads-
related operations.
 These primitives of the system are said to form a thread package.
 Some of the important issues in designing a thread package are:
1. Thread Creation
2. Thread Termination
3. Thread Synchronization
4. Thread Scheduling
5. Signal Handling

Thread Creation
 Threads can be created either statically or dynamically.
1. Static:
o The number of threads of a process remains fixed for its entire lifetime.
o Memory space is allocating to each thread.
2. Dynamic:
o The number of threads of a process keeps changing dynamically.
o Threads are created as and when it is needed during the process life cycle.
o It exits when task is completed.
o Here the stack size for the threads is specified as parameter to the system call
for thread creation.

Thread Termination
 Threads may terminate or never terminate until life cycle of process.
 Thread termination can be done as follows:
o Thread destroys itself on task completion by making an EXIT call.
o Thread is killed from outside using KILL command with thread identifier as parameter.
 In some process, all its threads are created immediately after the process start and then these
threads are never killed until the process terminates.

Thread Synchronization
 Since threads belongs to same process share the same address space, thread synchronization
is required to ensure that multiple threads don’t access the same data simultaneously.
 For example:
o If two threads want to increment the same global variable with in the same process.
o One thread should exclusive access to shared variable, increment it, and then pass
control to the other thread.
o It means that only one thread can execute in critical region at any instance of time.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 5

 Unit: 4 Processes and Processors
in Distributed Systems
Thread Scheduling
 Another important issue in designing threads package is to decide an appropriate scheduling
algorithm.
 Thread packages provide calls to give the users the flexibility to specify the scheduling policy
to be used for their applications.
 Some of the special features for threads scheduling that may be supported by a threads
package are as follows:
o Priority assignment facility
o Flexibility to vary quantum size dynamically
o Handoff scheduling
o Affinity scheduling

Signal Handling
 Signals provide software-generated interrupts and exceptions.
 Interrupts are externally generated disruptions of a thread or process.
 Exceptions are caused by the occurrence of unusual conditions during a thread's execution.
 The two main issues associated with handling signals in a multithreaded environment are as
follows:
o A signal must be handled properly no matter which thread of the process receives it.
o Signals must be prevented from getting lost when another signal of the same type
occurs in some other thread before the first one is handled by the thread in which it
occurred.

6) What is Scheduling in distributed systems? Explain Desirable features of a good Global

scheduling algorithm.
 A resource manager schedules the processes in a distributed system to make use of the system
resources.
 Scheduling is to optimize resource usage, response time, network congestion. It can be broadly
classified into three types:

 Task assignment approach

o In which each process is viewed as a collection of related tasks.
o These tasks are scheduled to suitable nodes so as to improve performance.

 Load-balancing approach
o In which all the processes submitted by the users are distributed among the nodes of
the system to equalize the workload among the nodes.

 Load-sharing approach
o Which simply attempts to conserve the ability of the system to perform work by
assuring that no node is idle while processes wait for being processed.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 6

 Unit: 4 Processes and Processors
in Distributed Systems
Desirable Features of a Good Global Scheduling Algorithm:

1. No A Priori knowledge about the Processes

 A good process scheduling algorithm should operate with absolutely no a priori knowledge
about the processes to be executed.
 Obtaining prior knowledge poses an extra burden upon the users who must specify this
information while submitting their processes for execution.

2. Dynamic in Nature
 Process assignment decisions should be based on the current load of the system and not on
some fixed static policy.
 Thus system support pre-emptive process migration facility in which a process can be migrated
from one node to another during the course of its execution.

3. Quick Decision making capability

 A good process scheduling algorithm must make quick decisions about the assignment of
processes to processors.
 For example, an algorithm that models the system by a mathematical program and solves it on
line is unsuitable because it does not meet this requirement.
 Heuristic methods requiring less computational effort while providing near optimal results are
therefore normally preferable to exhaustive solution methods.

4. Balance System Performance and Scheduling Overhead

 Several global scheduling algorithms collect global state information and use this information
in making process assignment decisions.
 A common intuition is that greater amounts of information describing global system state allow
more intelligent process assignment decisions to be made that have a positive effect on the
system as a whole.
 In a distributed environment, however, information regarding the state of the system is
typically gathered at a higher cost than in a centralized system.
 Hence algorithms that provide near optimal system performance with a minimum of global
state information gathering overhead are desirable.

5. Stability
 A scheduling algorithm is said to be unstable if it can enter a state in which all the nodes of the
system are spending all of their time migrating processes without accomplishing any useful
work in an attempt to properly schedule the processes for better performance.
 This form of fruitless migration of processes is known as processor thrashing.
 Processor thrashing can occur in situations where each node of the system has the power of
scheduling its own processes and scheduling decisions are based on relatively old data to
transmission delay between nodes.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 7

 Unit: 4 Processes and Processors
in Distributed Systems
 For example, it may happen that node n1 and n2 both observe that node n3 is idle and then
both offload a portion of their work to node n3 without being aware of the offloading decision
made by the other.
 Now if node n3 becomes overloaded due to the processes received from both nodes n1 and
n2, then it may again start transferring its processes to other nodes.
 This entire cycle may be repeated again and again, resulting in an unstable state.

6. Scalable
 Algorithm should be scalable and able to handle workload inquiry from any number of
machines in the network.
 The N2 nature of the algorithm creates more network traffic and quickly consumes network
bandwidth.
 A simple approach to make an algorithm scalable is to probe only m of N nodes for selecting a
host.
 The value of m can be dynamically adjusted depending upon the value of N.

7. Fault Tolerance.
 A good scheduling algorithm should not be disabled by the crash of one or more nodes of the
system.
 At any instance of time, it should continue functioning for nodes that are up at that time.

8. Fairness of Service.
 In any load balancing scheme, heavily loaded nodes will obtain all the benefits while tightly
loaded nodes will suffer poor response time A fair strategy that improves response time of
heavily loaded nodes without unduly affecting response time of poorly loaded node.

7) What is Processor Allocation OR Task Assignment in Distributed system? Explain Process

allocation OR Task assignment algorithms.

 Each process is divided into multiple tasks.

 These tasks are scheduled to suitable processor to improve performance.
 It requires characteristics of all the processes to be known in advance.
 This approach does not take into consideration the dynamically changing state of the system.
 In this approach, a process is considered to be composed of multiple tasks and the goal is to
find an optimal assignment policy for the tasks of an individual process.

Task Assignment Approach algorithms

1. Graph Theoretic Deterministic Algorithm

 A system with m CPUs and n processes has any of the following three cases:
 m=n: Each process is allocated to one CPU.
 M>n: Some CPUs may remain idle (free) or work on earlier allocated processes.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 8

 Unit: 4 Processes and Processors
in Distributed Systems
 M<n: There is a need to schedule processes on CPUs, and several processes may be assigned
to each CPU.
 The main objective of performing CPU assignment is to:
o Minimize IPC cost.
o Obtain quick turnaround time.
o Achieve high degree of parallelism for efficient utilization.
o Minimize network traffic.

 Processes are represented as nodes A, B, C, D, E, F, G and H.

 Arcs between sub-graphs represent network traffic and their weights represent IPC costs.
 Total network traffic is the sum of the arcs intersected by the dotted cut lines.

 Partition 1
o CPU 1 runs A, D, G
o CPU 2 runs B, E, F, H and C
o Network traffic= 4+3+2+5=14
 Partition 2
o CPU 1 runs processes A, D, E, G, B
o CPU2 runs processes H, C and F
o Network traffic= 5+2+4=11
 Thus partition 2 communication generates less network traffic as compared to partition 1.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 9

 Unit: 4 Processes and Processors
in Distributed Systems
Centralized Heuristic Algorithm

 It is also called Top down algorithm.

 It doesn’t require advance information.
 Coordinator maintains the usage table with one entry for every user (processor) and this is
initially zero.
 Usage table entries can either be zero, positive, or negative.
o Zero value indicates a neutral state.
o Positive value implies that the machine is using system resources.
o Negative value means that the machine needs resources.

Hierarchical Algorithm

 Process hierarchy is modelled like an organization hierarchy.

 For each group of workers, one manager machine (department head) is assigned the task of
keeping track of who is busy and who is idle.
 If the system is large, there will be number of department heads, so some machines will
function as “deans”.
 Each processor maintains communication with one superior and few subordinates.
 When a dean or department head stops functioning (crashes), promote one of the direct
subordinates of the faulty manager to fill in for the boss.
 The choice of which can be made by the subordinates themselves.
Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 10
 Unit: 4 Processes and Processors
in Distributed Systems
8) Classify Load balancing approaches and Explain each in detail. Write advantages of Load
balancing.
 The distribution of loads to the processing elements is simply called the load balancing.
 Load balancing approaches can be classified as follows:

Figure: Classification of Load Balancing approaches

Static Load Balancing
 In static algorithm the processes are assigned to the processors at the compile time according
to the performance of the nodes.
 Once the processes are assigned, no change or reassignment is possible at the run time.
 Number of jobs in each node is fixed in static load balancing algorithm.
 Static algorithms do not collect any information about the nodes.
 The static load balancing algorithms can be divided into two sub classes:
1. Optimal Static Load Balancing Algorithm
o If all the information and resources related to a system are known optimal static
load balancing can be done.
2. Sub optimal static load balancing Algorithm
o Sub-optimal load balancing algorithm will be mandatory for some applications
when optimal solution is not found.

Dynamic Load Balancing

 In dynamic load balancing algorithm assignment of jobs is done at the runtime.
 In DLB jobs are reassigned at the runtime depending upon the situation.
 The load will be transferred from heavily loaded nodes to the lightly loaded nodes.
 No decision is taken until the process gets executed.
 This strategy collects the information about the system state and about the job information.
 As more information is collected, the algorithm can make better decision.
Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 11
 Unit: 4 Processes and Processors
in Distributed Systems
Deterministic vs Probabilistic
 Deterministic algorithms are suitable when the process behavior is known in advance.
 If all the details like list of processes, computing requirements, file requirements and
communication requirements are known prior to execution, then it is possible to make a
perfect assignment.
 In the case load is unpredictable or variable from minute to minute or hour to hour, a
Probabilistic/heuristic processor allocation is preferred.

Centralized Vs Distributed
CENTRALIZED LOAD BALANCING DISTRIBUTED LOAD BALANCING
Scheduling decision is carried out at one single Scheduling decision is carried out at different
node called the centralized node. nodes called the Distributed nodes.
Every information is available at a single node. Information is distributed at different nodes.
Centralized approach leads to a bottleneck as Distributed approach handle the multiple
number of requests increases. requests by physically distributing among
various nodes.

If centralized server fails, all scheduling in the If distributed server fails, all scheduling will be
system will fails. handled by other servers.

Advantages of Load balancing

 Load balancing improves the performance of each node and hence the overall system
performance will improve.
 Load balancing reduces the job idle time.
 Low cost but high gain.
 Small jobs do not suffer from long starvation.
 Extensibility and incremental growth.
 Maximum utilization of resources.
 Higher throughput.
 Response time becomes shorter.

9) Explain issues in Designing Load Balancing Algorithms.

 Load estimation: determines how to estimate the workload of a node in a distributed system.
 Process transfer: decides whether the process can be executed locally or remotely.
 Static information exchange: determines how the system load information can be exchanged
among the nodes.
 Location policy: determines the selection of a destination node during process migration.
 Priority assignment: determines the priority of execution of a set of local and remote processes
on a particular node.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 12

 Unit: 4 Processes and Processors
in Distributed Systems
 Migration limiting policy: determines the total number of times a process can migrate from
one node to another.

10) What is Load sharing in Distributed system? Explain Location policies in Designing Load Sharing
Algorithms.
 Load sharing algorithms ensure that no node is idle or heavily loaded.
 Policies for load sharing approach are the same as load balancing polices.
 They include load estimation policy, process transfer policy, location policy and state
information exchange.
 They differ in location policy.

Location Policies
 The location policy decides the sender node or the receiver node of a process that is to be
moved within the system for load sharing.
 Depending on the type of node that takes the initiative to globally search for a suitable node
for the process, the location policies are of the following types:

1. Sender-initiated policy:
o In which the sender node of the process decides where to send the process.
o The heavily loaded nodes search for lightly loaded nodes to which work may be
transferred.
o when a node's load becomes more than the threshold value, it either broadcasts a
message or randomly probes the other nodes one by one to find a lightly loaded node
that can accept one or more of its processes.
o If a suitable receiver node is not found, the node on which the process originated must
execute that process.

2. Receiver-initiated policy:
o In which the receiver node of the process decides from where to get the process.
o In this policy lightly loaded nodes search for heavily loaded nodes from which processes
can be accepted for execution.
o When the load on a node falls below a threshold value, it broadcasts a probe message
to all nodes or probes nodes one by one to search for a heavily loaded node.
o Some heavily loaded node may transfer one of its process if such a transfer does not
reduce its load below normal threshold.

11) What is process migration? List the steps for it. What are the advantages of process migration?
 Process migration is a specialized form of process management whereby processes are moved
from one computing environment to another.
 Process migration is classified into two types:
o Non-preemptive: Process is migrated before execution start in source node.
o Preemptive: Process is migrated during its execution.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 13

 Unit: 4 Processes and Processors
in Distributed Systems

Figure: Flow of execution of a migrating process

 Process migration involves the following major steps:

1. Selection of a process that should be migrated.
2. Selection of the destination node to which the selected process should be migrated.
3. Actual transfer of the selected process to the destination node.

Advantages of Process Migration

1. Reducing average response time of processors

o Process migration facility is being used to reduce the average response time of the
processes of a heavily loaded node by some processes on idle or underutilized nodes.

2. Speeding up individual jobs

o Process migration facility may be used to speed up individual jobs in two ways.
o The first method is to migrate the tasks of a job to the different nodes of the system
and to execute them concurrently.
o The second approach is to migrate a job to a node having a faster CPU or to a node at
which it has minimum turnaround time.

3. Gaining higher throughout

o In a system with process migration facility, the capabilities of the CPUs of all the nodes
can be better utilized by using a suitable load-balancing policy.
o This helps in improving the throughput of the system.

4. Utilizing resources effectively

o Process migration policy also helps in utilizing resources effectively, since any
distributed system consisting of different resources such as CPU, printers, storage etc.
and software (databases, files with different capabilities are optimally used.
o Depending nature of process, it can be appropriately migrated to utilize the system
resource efficiently.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 14

 Unit: 4 Processes and Processors
in Distributed Systems
5. Reducing network traffic
o Migrating a process closer to the resource it is using most heavily may reduce network
traffic in the system if the decreased cost of accessing its favorite resources offsets the
possible increased cost of accessing its less favored ones.
o Another way to reduce network traffic by process migration is to migrate and cluster
two or more processes, which frequently communicate with each other, on the same
node of the system.

6. Improving system reliability

o Simply migrate a critical process to a node whose reliability is higher than other nodes
in the system.
o Migrate a copy of a critical process to some other node and to execute both the original
and copied processes concurrently on different nodes.
o In failure modes such as manual shutdown, which manifest themselves as gradual
degradation of a node, the process of the node, for their continued execution, may be
migrated to another node before the dying node completely fails.

7. Improving system security

o A sensitive process may be migrated and run on a secure node that is not directly
accessible to general users, thus improving the security of that process.

12) What are the desirable features of a good process migration mechanism?

Desirable Features of a Good process migration mechanism:

1. Transparency
 Transparency is an important requirement for a system that supports process migration.
 Object access level transparency
o Object access level transparency is the minimum requirement for a system to support
non preemptive process migration
o Access to objects such as files and devices can be done in a location independent
manner.
o The object access level transparency allows, free initiation of programs at an arbitrary
node.
 System call and inter process communication transparency
o System call and inter process communication level transparency is must for a system to
support preemptive process migration facility.
o Transparency of inter process communication is also desired for the transparent
redirection of messages during the transient state of process that recently migrated.

2. Minimal Interface
 Migration of a process should cause minimal interference to the progress of the process
involved and to the system as a whole.
 This can be achieved by minimizing the freezing time of the process being migrated.
Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 15
 Unit: 4 Processes and Processors
in Distributed Systems
3. Minimal Residual Dependencies
 A migrated process should not in any way continue to depend on its previous node once it has
started executing on its new node since, otherwise, the following will occur.

4. Efficiency
 It is the major issue in implementing process migration.
 Main sources of inefficiency are: time required to migrate a process, the cost of locating an
object and the code of supporting remote execution once the process is migrated.

5. Robustness
 The failure of a node other than the one on which a process is currently running should not in
any way affect the accessibility or execution of that process.

6. Communication between coprocesses of a job

 Benefit of process migration is the parallel processing among the processes of a single job
distributed over several nodes.
 Coprocesses are able to directly communicate with each other irrespective of their locations.

13) What is Fault Tolerance. Explain classification of Component faults. Explain technique to handle
Redundancy.
 The main objective of designing a fault tolerance system is to ensure that the system operates
correctly even in presence of faults.
 Distributed system made up of independent computers connected over the network, can fail
due to fault in any of hardware components or software components.
 Component faults can be classified as:

1. Transient faults
 It can occur suddenly, disappear, and may not occur again if the operation is repeated.
 For example:
o During heavy traffic in network, a telephone call is misroute but if call is retried, it will
reach destination correctly.

2. Intermittent faults
 These faults occur often, but may not be periodic in nature.
 For example:
o The loose connection to network switch may cause intermittent connection problems.
o Such faults are difficult to diagnose.
o During debugging process, it is possible that the fault may not be detected at all.
3. Permanent faults
 These faults can be easily identified and the components can be replaced.
 For example:
o Software bug or disk head crash causes permanent faults.
o These faults can be identified and corresponding action can be taken.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 16

 Unit: 4 Processes and Processors
in Distributed Systems
 Technique to handle fault tolerance is Redundancy.
 Types of Redundancy are as follows:

1. Information redundancy
 Extra bits are added to data to handle fault tolerance by detecting errors.

2. Time redundancy
 An action performed once is repeated if needed after a specific time period.
 For example, if an atomic transaction aborts, it can be executed without any side effects.
 Time redundancy is helpful when the faults are transient or intermittent.

3. Physical redundancy
 Extra equipment's are added to enable the system to tolerate faults due to loss or malfunction
of some components.
 For example, extra stand by processors can be used in the system.

14) What is Real time distributed system? Explain the types of Real time distributed system.
 It is developed by collection of computers in a network.
 Some of these are connected to external devices that produce or accept data or expect to be
controlled in real time.
 Computers may be tiny microcontrollers or stand-alone machines.
 In both cases they usually have sensors for receiving signals from the devices and actuators for
sending signals to them.
 The sensors and actuators may be digital or analog.

Figure: A Distributed real-time computer system

 Real-time systems are generally split into two types depending on how serious their deadlines
are and the consequences of missing one.
1. Soft real time system: missing an occasional deadline is all right. For example, a telephone
switch might be permitted to lose or misroute one call in 105 under overload conditions and
still be within specification.
2. Hard real time system: even a single missed deadline in a hard real-time system is
unacceptable, as this might lead to loss of life or an environmental catastrophe

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 17

 Unit: 5 Distributed File System
1) What is Distributed file system? Explain Benefits and Services of distributed file system.
 A file is a subsystem of an operating system that performs file management activities such as
organization, storing, retrieval, naming, sharing and protection of files.
 A distributed file system (DFS) is a file system with data stored on a server.
 The data is accessed and processed as if it was stored on the local client machine.
 The DFS makes it convenient to share information and files among users on a network in a
controlled and authorized way.
 Server allows the client users to share files and store data just like they are storing the
information locally.
 Servers have full control over the data and give access control to the clients.

Benefits of distributed file system

1. Remote Information sharing
 A distributed file system allows a file to be transparently accessed by processes
of any
 node of the system irrespective of file’s location.
2. User mobility
 User should not be forced to work on one specific node but should have
flexibility to work on different nodes at different times.
3. Availability
 For better fault tolerance file should be available for use even in the event of
temporary failure of one or more nodes of the system.
4. Diskless workstation
 A diskless workstation is more economical, less noisy and generates less heat.
 A distributed system with its transparent remote file sharing capability allows
use of diskless workstation in system.

Services of distributed file system

1. Storage service
 Deals with allocation and management of secondary device spaces.
 It provides a logical view of the storage system by providing operations for
storing and retrieving data.
 It is also known as disk service and block service.
2. True file service.
 It is concerned with the operations on individual file such as operation for
accessing and modifying the data in files and for creating and deleting files.
3. Name/Directory service
 It provides mapping between text names for files to their IDs.
 It performs directory related activities such as creation and deletion of
directories, adding new file to directory, deleting file from directory, changing
name of file, moving a file from one directory to another.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 1

 Unit: 5 Distributed File System
2) Explain feature and goals of distributed file system.
1. Transparency
 Structure transparency
o Clients should not know the number or locations of file servers and the storage devices.
 Access transparency
o Both local and remote files should be accessible in the same way.
o The file system should automatically locate an accessed files and arrange for the
transport of data to the client's site.
 Naming transparency
o The name of a file should give no hint as to where the file is located.
o The name of the file must not be changed when moving from one node to another.
 Replication transparency
o If a file is replicated on multiple nodes, both the existence of multiple copies and their
locations should be hidden from the clients.

2. User mobility
 The user should not be forced to work on a specific node but should have the flexibility to work
on different nodes at different times.
 This can be achieved by automatically bringing the users environment to the node where the
user logs in.

3. Performance
 Performance is measured as the average amount of time needed to satisfy client requests.
 This time includes CPU time + time for accessing secondary storage + network access time.
 It is desirable that the performance of a distributed file system should be comparable to that
of a centralized file system.

4. Simplicity and ease of use

 User interface to the file system must be simple and the number of commands should be as
small as possible.

5. Scalability
 A good distributed file system should be designed to easily cope with the growth of nodes and
users in the system.
 Such growth should not cause serious disruption of service or significant loss of performance
to users.

6. High availability
 A distributed file system should continue to function even when partial failures occur due to
the failure of one or more components.
 It should have multiple and independent file servers controlling multiple and independent
storage devices.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 2

 Unit: 5 Distributed File System
7. High reliability
 The probability of loss of stored data should be minimized.
 System should automatically generate backup copies of critical files that can be used in the
event of loss of the original ones.

8. Data integrity
 Concurrent access requests from multiple users must be properly synchronized by the use of
some concurrency control mechanism.
 Atomic transactions are a high-level concurrency control mechanism provided to users by a file
system for data integrity.

9. Security
 Users should be confident about the privacy of their data.
 Necessary security mechanisms must be implemented against unauthorized access of files.

10. Heterogeneity
 There should be easy access to shared data on diverse platforms (e.g. Unix workstation, Wintel
platform etc.).

3) Explain File models of a distributed file system.

Figure: Classification of File models

Unstructured files
 In this model, there is no substructure known to the file server.
 Contents of each file of the file system appears to the file server as an uninterpreted sequence
of bytes.
 Interpretation of the meaning and structure of the data stored in the files are entirely up to the
application programs.
 UNIX, MS-DOS and other modern operating systems use this file model.
 This is mainly because sharing of a file by different applications is easier compared to the
structured file model.
Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 3
 Unit: 5 Distributed File System
Structured files
 In structured files (rarely used now) a file appears to the file server as an ordered sequence of
records.
 Records of different files of the same file system can be of different sizes.
 NetBSD operating system use this file model.
 Two types of structured files are:
1. Files with non-indexed records
o File record is accessed by specifying its position within the file.
o For example, the fifth record from the beginning of the file or the second record from
the end of the file.
2. Files with indexed records
o Records have one or more key fields and can be addressed by specifying the values of
the key fields.
o File is maintained as B-tree or other suitable data structure or hash table to locate
records quickly.

Mutable and Immutable files

MUTABLE FILES IMMUTABLE FILES
Update performed on a file overwrites on its Rather than updating the same file, a new
old contents to produce the new contents. version of the file is created each time a change
is made to the file contents.

A file can be modified by each update A file cannot be modified once it has been
operation. created except to be deleted.
File overwrites its old contents to produce the File versioning approach is used to implement
new contents. file updates.
Most existing operating systems use the It is rarely used now a day.
mutable file model.

4) Explain File accessing models of a distributed file system.

 The file accessing model of a distributed file system mainly depends on two factors:
1. The method used for accessing remote files
2. The unit of data access

1. Accessing Remote Files

 A distributed file system may use one of the following models to service client’s file access
request.
 Remote service model
o The client's request for file access is delivered to the server, the server machine
performs the access request, and finally the result is forwarded back to the client.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 4

 Unit: 5 Distributed File System
o The access requests from the client and the server replies for the client are transferred
across the network as messages.
o The file server interface and the communication protocols must be designed carefully
to minimize the overhead of generating messages as well as the number of messages
that must be exchanged in order to satisfy a particular request.
 Data caching model
o If the data needed to satisfy the client's access request is not present locally, it is copied
from the server’s node to the client's node and is cached there.
o The client's request is processed on the client's node itself by using the cached data.
o Thus repeated accesses to the same data can be handled locally.
o A replacement strategy LRU is used is used for cache management.

2. Unit of Data Transfer

 Unit of data refers to fraction of file data that is transferred to and from clients as a result of a
single read or write operations.
 File level transfer model
o Whole file is treated as unit of data transfer between client and server in both the
direction.
Advantages
o Transmitting an entire file in response to a single request is more efficient than
transmitting it page by page as the network protocol overhead is required only once.
o It has better scalability because it requires fewer accesses to file servers, resulting in
reduced server load and network traffic.
o Disk access routines on the servers can be better optimized if it is known that requests
are always for entire files rather than for random disk blocks.
o Once an entire file is cached at a client's site, it becomes immune to server and network
failures.
Disadvantage
o This model requires sufficient storage space on the client's node for storing all the
requires files in their entirely.

 Block level transfer model

o In this model, file data transfer across the network between a client and a server take
place in units of file blocks.
o A file block is a contiguous portion of a file and is usually fixed in length.
o In page level transfer model block size is equal to virtual memory page size.
Advantages
o This model does not require client nodes to have large storage.
o It eliminates the need to copy an entire file when only a small portion of the file data is
needed.
o It provides large virtual memory for client nodes that do not have their own secondary
storage devices.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 5

 Unit: 5 Distributed File System
Disadvantage
o When an entire file is to be accessed, multiple server requests are needed in this model,
results in more network traffic and more network protocol overhead

 Byte level transfer model

o In this model, file data transfers across the network between a client and a server take
pace in units of bytes.
Advantages
o This model provides maximum provides maximum flexibility because it allows storage
and retrieval of an arbitrary sequential subrange of a file, specified by an offset within
a file, and a length.
o Cache management is difficult due to variable length data for different requests.

 Record level transfer model

o In this model, file data transfers across the network between a client and a server take
place in units of records.
5) Explain File sharing semantics in distributed file system.
 A shared file may be simultaneously accessed by multiple users.
 In such a situation, an important design issue for any file system is to clearly define when
modifications of file data made by a user are observable by other users.

1. Unix semantics
 This semantics enforces an absolute time ordering on all operations.
 Every read operation on a file sees the effects of all previous write operations performed on
that file.
 This semantics can be achieved in a distributed system by disallowing files to be cached at client
nodes.
 Allowing a shared file to be managed by only one file server that processes all read and write
requests for the file strictly in the order in which it receives them.
 There is a possibility that, due to network delays, client requests from different nodes may
arrive and get processed at the server node out of order.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 6

 Unit: 5 Distributed File System

Figure: Unix semantics

2. Session semantic
 A client opens a file, performs a series of read/write operations on the file, and finally closes
the file when he or she is done with the file.
 A session is a series of file accesses made between the open and close operations.
 In session semantics, all changes made to a file during a session are initially made visible only
to the client process that opened the session and are invisible to other remote processes who
have the same file open simultaneously.
 Once the session is closed, the changes made to the file are made visible to remote processes
only in later starting sessions.

Figure: Session Semantics

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 7

 Unit: 5 Distributed File System
3. Immutable shared file semantics
 According to this semantics, once the creator of a file declares it to be sharable, the file is
treated as immutable, so that it cannot be modified any more.
 Change to the file are handled by creating a new updated version of the file.
 Therefore, the semantics allows files to be shared only in the read-only mode.

Figure: Immutable File semantics

4. Transaction like semantics
 This semantics is based on the transaction mechanism, which is a high-level mechanism for
controlling concurrent access to shared mutable data.
 A transaction is a set of operations enclosed in-between a pair of begin_transaction and
end_transaction like operations.
 The transaction mechanism ensures that the partial modifications made to the shared data by
a transaction will not be visible to other concurrently executing transactions until the
transaction ends.

6) What is File caching scheme? Explain key decisions of File caching in Distributed File System.
 The idea in file caching is to retain recently accessed file data in main memory, to avoid
repeated accesses.
 File caching reduces disk transfers substantially, resulting in better overall performance of the
file system.
 A file-caching scheme for a distributed file system should address the following key decisions:
o Cache location
o Modification propagation
o Cache validation
1. Cache location
 Cache location refers to the place where the cached data is stored.
 Assuming that the original location of a file is on its server's disk, there are these possible cache
locations in a distributed file system.
 Server’s Main Memory
o When no caching scheme is used, before a remote client can access a file, the file must
first be transferred from the server's disk to the server's main memory.
o Then across the network from the server's main memory to the client's main memory.
Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 8
 Unit: 5 Distributed File System
o Thus the total cost involved is one disk access and one network access.
Advantages
o It is easy to implement and is totally transparent to the clients.
o It is easy to keep the original file and cached data consistent.
o Since a single server manages both the cached data and the file, multiple accesses from
different clients can be easily synchronized to support UNIX-like file-sharing semantics.
Disadvantages
o It does not eliminate the network access cost and does not contribute to the scalability
and reliability of the distributed file system.

 Client’s Disk
 A cache located in a client's disk eliminates network access cost but requires disk access cost
on a cache hit.
Advantages
o If the cached data is kept on the client's disk, the data is still there during recovery and
there is no need to fetch it again from the server's node.
o As compared to a main-memory cache, a disk cache has plenty of storage space.
o Disk cache is useful to those applications that use file level transfer.
Disadvantages
o This policy does not work if the system is to support diskless workstations.
o Access time increases as every time disk access is done.

 Client main memory

 A cache located in a client's main memory eliminates both network access cost and disk access
cost.
 It also permits workstations to be diskless.
Advantages
o It provides high scalability and reliability as cache hit access request can be serviced
locally without the need to contact server.
Disadvantages
o Client disk memory is small compared to client’s disk size.

2. Modification propagation
 Write through Scheme
 In this scheme, when a cache entry is modified, the new value is immediately sent to the server
for updating the master copy of the file.
Advantages
o High degree of reliability and suitability for UNIX-like semantics
o Since every modification is immediately propagated to the server having the master
copy of the file, the risk of updated data getting lost is very low.
Disadvantages
o It has poor write performance as write access has to wait until the information is written
to the master copy of the server.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 9

 Unit: 5 Distributed File System
o This scheme is suitable for use only in those cases in which the ratio of read-to-write
accesses is fairly large.

 Delayed Write
 In this scheme, when a cache entry is modified, the new value is written only to the cache and
the client just makes a note that the cache entry has been updated.
 After some time, all updated cache entries corresponding to a file are gathered together and
sent to the server at a time.
 Depending on when the modifications are sent to the file server, delayed-write policies are of
different types.
Advantages
o Write accesses complete more quickly because the new value is written only in the
cache of the client performing the write.
o Modified data may be deleted before it is time to send them to the server, in such cases,
modifications need not be propagated at all to the serve, resulting in a major
performance gain.
o Gathering of all file updated and sending them together to the server is more efficient
than sending each update separately.

3. Cache Validation Schemes

 A client's cache entry becomes stale as soon as some other client modifies the data
corresponding to the cache entry in the master copy of the file.
 Therefore, it becomes necessary to verify if the data cached at a client node is consistent with
the master copy.
 If not, the cached data must be invalidated and the updated version of the data must be fetched
again from the server.
 There are two approaches:

a) Client Initiated Approach

o In this approach, a client contacts the server and checks whether its locally cached data
is consistent with the master copy.
o Checking before every access
 This approach defeats the main purpose of caching because the server has to be
contacted on every access.
 It is suitable for supporting UNIX-like semantics.
o Periodic checking
 In this method, a check is initiated every fixed interval of time.
o Check on file open
 In this method, a client's cache entry is validated only when the client opens the
corresponding file for use.
 This method is suitable for supporting session semantics.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 10

 Unit: 5 Distributed File System
b) Server Initiated Approach
o If the frequency of the validity check is high, the client-initiated cache validation
approach generates a large amount of network traffic and consumes precious server
CPU time.
o In this method, a client informs the file server when opening a file, indicating whether
the file is being opened for reading, writing, or both.
o The file server keeps a record of which client has which file open and in what mode.
o The server keeps monitoring the file usage modes being used by different clients and
reacts whenever it detects a potential for inconsistency.
o When a new client makes a request to open an already open file and if the server finds
that the new open mode conflicts with the already open mode.
o The server can deny the request or queue the request or disable caching and switch to
the remote service mode of operation for that particular file by asking all the clients
having the file opens to remove that file from their caches.
Disadvantages
o It violates the traditional client-server model in which servers simply respond to service
request activities by clients.
o It requires that file servers be stateful.
o Client – initiated cache validation, approach must still be used along with the server
initiated approach.
o For example, a client may open a file, cache it, and then close it after use.
o Upon opening it again for use, the cache content must be validated because there is a
possibility that some other client might have subsequently opened, modified and closed
the file.
7) What is File replication? Write the Advantages of File Replication.
 File replication is the primary mechanism for improving file availability.
 A replicated file is a file that has multiple copies, with each copy located on a separate file
server.
 The main problem in Replication is consistency, i.e. when one copy changes, how do other
copies reflect that change?
 Often there is a tradeoff: consistency versus availability and performance.
Advantages of Replication
1. Increased availability
o The system remains operational and available to the users despite failures.
o By replicating critical data on servers with independent failure modes, the probability
that one copies of the data will be accessible increases.
2. Increased reliability
o Replication allows the existence of multiple copies of their files.
o Due to the presence of redundant information in the system, recovery from failures
becomes possible.
3. Improved response time
o It enables data to be accessed either locally or from a node to which access time is lower
than the primary copy access time.
Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 11
 Unit: 5 Distributed File System
4. Reduced network traffic
o If a file's replica is available with a file server that resides on a client's node, the client's
access requests can be serviced locally, resulting in reduced network traffic.
5. Improved system throughput
o Replication also enables several client’s requests for access to the same file to be
serviced in parallel by different servers, resulting in improved system throughput.
6. Better scalability
o By replicating the file on multiple servers, the same requests can now be serviced more
efficiently by multiple servers due to workload distribution.
o This results in better scalability.
7. Autonomous operation
o In a distributed system that provides file replication as a service to their clients, all files
required by a client for operating during a limited time period may be replicated on the
file server residing at the client's node.
o This will facilitate temporary autonomous operation of client machines.
8) Give difference between replication and caching.

REPLICATION CACHING
Replica is associated with a server. Cached copy is normally associated with a
client.
Existence of a replica depends on availability Existence of a cached copy is dependent on
and performance requirement. the locality in file access patterns.
More persistent. Less persistent.
Widely known, secure, available, complete Less known, secure, available, complete, and
and accurate. accurate.
A replica does not change upon cached. A cached copy is subject to change upon a
replica.

9) Write a note on Replication Transparency.

 Replication of files should be designed to be transparent to the users so that multiple copies of
a replicated file appear as a single logical file to its users.
1. Naming of Replicas
o Assignment of a single identifier to all replicas of an object seems reasonable for
immutable objects because a kernel can easily support this type of object.
o All copies are immutable and identical; there is only one logical object with a give
identifier
o However, in mutable objects, different copies of a replicated object may not be the
same (consistent) at a particular instance of time.
o In this case, if the same identifier is used for all replicas of the object, the kernel cannot
decide which replica is the most up-to-date one.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 12

 Unit: 5 Distributed File System
o Therefore, the consistency control and management of the various replicas of a mutable
object should be performed outside the kernel.
o It is the responsibility of the naming system to map a user-supplied identifier into the
appropriate replica of a mutable object.
2. Replication control
o Replication control includes determining the number and locations of replicas of a
replicated file.
o Depending on whether replication control is user transparent or not, the replication
process is of two types:

a) Explicit replication
o In this type, users are given the flexibility to control the entire replication process.
o That is, when a process creates a file, it specifies the server on which the file should be
placed.
o Then, if desired, additional copies of the file can be created on other servers on explicit
request by the users.
o Users also have the flexibility to delete one or more replicas of a replicated file.

b) Implicit/lazy replication
o In this type, the entire replication process is automatically controlled by the system
without user’s knowledge.
o That is, when a process creates a file, it does not provide any information about its
location.
o The system automatically selects one server for the placement of the file.
o Later the system automatically creates replicas of the file on other servers, based on
some replication policy used by the system.

10) Explain Multicopy Update Problem in File Replication and Approaches to handle it.
 As soon as a file system allows multiple copies of the same (logical) file to exist on different
servers, it is faced with the problem of keeping them mutually consistent.

1. Read Only Replication

 This approach allows the replication of only immutable files.
 Files known to be frequently read and modified only once in a while such as files containing the
object code of system programs, can be treated as immutable files for replication using this
approach.

2. Read Any Write All Protocol

 In this method a read operation on a replicated file is performed by reading any copy of the file
and a write operation by writing to all copies of the file.
 Before updating any copy, all copies are locked, then they are updated, and finally the locks are
released to complete the write.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 13

 Unit: 5 Distributed File System
3. Available Copies Protocol
 This method allows write operations to be carried out even when some of the servers having a
copy of the replicated file are down.
 In this method, a read operation is performed by reading any available copy, but a write
operation is performed by writing to all available copies.
 When a server recovers after a failure, it brings itself up to date by copying from other servers
before accepting any user request.
 Failed servers are dynamically detected by high-priority status management routines and
configured out of the system while newly recovered sites are configured back in.

4. Primary Copy Protocol

 In this protocol, for each replicated file, one copy is designated as the primary copy and all the
others are secondary copies.
 Read operations can be performed using any copy, primary or secondary.
 But all write operations are directly performed only on the primary copy.
 Each server having a secondary copy updates its copy either by receiving notification of changes
from the server having the primary copy or by requesting the updated copy from it.

5. Quorum-Based Protocols
 Suppose that there are a total of n copies of a replicated file F.
 To read the file, a minimum r copies of F have to be consulted.
 This set of r copies is called a read quorum.
 To perform a write operation on the file, a minimum w copies of F have to be written.
 This set of w copies is called a write quorum.
 The restriction on the choice of the values of r and w is (r+w>n).
 That is, there is at least one common copy of the file between every pair of read and write
operations resulting in at least one up-to-date copy in any read/write quorum.
 The version number of a copy is updated every time the copy is modified.
 A copy with the largest version number in a quorum is current.
 A read is executed as follows:
o Retrieve a read quorum (any r copies) of F.
o Of the r copies retrieved, select the copy with the largest version number.
o Perform the read operation on the selected copy.
 A write is executed as follows:
o Retrieve a write quorum (any w copies) of F.
o Of the w copies retrieved, get the version number of the copy with the largest version
number.
o Increment the version number.
o Write the new value and the new version number to all the w copies of the write
quorum.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 14

 Unit: 5 Distributed File System
 Examples of quorum consensus algorithm:(n=8, r=4, w=5)

Figure: Example of Quorum based protocol

o There are total of eight copies of the replicated file (n = 8).
o The values of read and write quorums are 4 and 5, respectively (r =4, w= 5).
o The condition r +w > n is satisfied.
o Now suppose a write operation is performed on the write quorum comprised of copies
3, 4, 5, 6, and 8.
o All these copies get the new version and the new version number.
o Now any subsequent read operation will require a read quorum of four copies because
r=4.
o any read quorum will have to contain at least one copy of the previous write quorum.

11) Explain file and Storage properties that directly influence the ability of a distributed file system
to tolerate faults.
 The primary file properties that directly influence the ability of a distributed file system to
tolerate faults are as follow:

1. Availability
o Availability of a file refers to the fraction of time for which the file is available for use.
o If a network is partitioned due to a communication link failure, a file may be available
to the clients of some nodes, but at the same time, it may not be available to the clients
of other nodes.

2. Robustness
o Robustness of a file refers to its power to survive crashes of the storage device and
decays of the storage medium on which it is stored.
o Storage devices that are implemented by using redundancy techniques, such as a stable
storage device, are often used to store robust files.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 15

 Unit: 5 Distributed File System
3. Recoverability
o Recoverability of file refers to its ability to be rolled back to an earlier, consistent state
when an operation on the file fails or is aborted by the client.

 In context of crash resistance capability, storage may be broadly classified into three types:
1. Volatile Storage (RAM)
o It cannot withstand power failure or machine crash.
o Data is lost in case of power failure or machine crash.

2. Nonvolatile storage (Disk)

o It can withstand CPU failures but cannot withstand transient I/O faults and decay of the
storage media.

3. Stable Storage
o It can withstand transient I/O fault and decay of storage media.

12) Explain Effect of Service Paradigm on Fault Tolerance.

 A server may be implemented by using any one of the following two service paradigms stateful
and stateless.
1. Stateful File Servers
o A stateful server maintains client’s state information from one remote procedure call
to the next.
o In case of two subsequent calls by a client to a stateful server, some state information
pertaining to the service performed for the client as a result of the first call execution is
stored by the server process.
o These clients state information is subsequently used at the time of executing the second
call.

Figure: Stateful File Server

o For example, let us consider a server for byte-stream files that allows the following
operations on files.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 16

 Unit: 5 Distributed File System
o Following Figure shows example of stateful server.
1. Open(filename,mode)
 This operation is used to open a file identified by filename in the specified mode.
 When the server executes this operation, it creates an entry for this file in file
table that it uses for maintaining the file state information of all the open files.
 The file state information normally consists of the identifier of the file, the open
mode, and the current position of a nonnegative integer pointer, called the read
writer pointer.
 When a file is opened, it read-write pointer is set to zero and the server returns
to the client a file identifier(fid), which is used by the client for subsequent
accesses to that file.

2. Read(fid,n,buffer)
 This operation is used to get n bytes of data from the file identified by fid into
the buffer named buffer.
 When the server executes this operation, it returns to the client n bytes of file
data starting from the byte currently addressed by the read write pointer and
then increments the read writer pointer by n.

3. Write(fid,n,buffer)
 On execution of this operation, the server takes n bytes of data from the
specified buffer, writes it into file identified by fid at the byte position currently
addressed by the read write pointer and then increments the read write pointer
by n.

4. Seek(fid,position)
 This operation causes the server to change the value of the read-write pointer
of the file identified by fid to the new value specified as position.

5. Close(fid)
 This statement causes the server to delete from its file table the file state
information of the file identified by fid.

2. Stateless File server

 A stateless server does not maintain any client state information.
 Every request from client is accompanied with all the necessary parameters to successfully
carry out desired operation.
 For example, a server for byte stream files that allow the following operations on file is
stateless.
1. Read(filename,position,n,buffer)
 On execution of this operation, the server returns to the client n bytes of data
of the file identified by filename.
 The returned data is placed in the buffer named buffer.
 The value of actual number of bytes read is also returned to the client.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 17

 Unit: 5 Distributed File System
 The position within the file from where to being reading is specified as the
position parameter.

2. Write(filename,position,n,buffer)
 When the server executes this operation, it takes n bytes of data from the
specified buffer and writes it into the file identified by filename.
 The position parameter specifies the byte position within the file from where to
start writing.
 The server returns to the client the actual number of bytes written.
 If a client wishes to have similar effect as above Figure, the following two read operation must
be carried out:
o Read(filename,0,100,buf)
o Read(filename,100,100,buf)

Figure: Stateless File Server

13) Give difference between Stateful File Server and Stateless File server.
PARAMETERS STATEFUL SERVER STATELESS SERVER
State A Stateful server remember client A Stateless server does not remember
data (state) from one request to the state information.
next.
Programming Stateful server is harder to code. Stateless server is straightforward to
code.
Efficiency More because clients do not have to Less because information needs to be
provide full file information every provided.
time they perform an operation.
Crash Difficult due to loss of information. Can easily recover from failure because
recovery there is no state that must be restored.
Information The client can send less data with The client must specify complete file
transfer each request. names in each request.
Operations Open, Read, Write, Seek, Close Read, Write

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 18

 Unit: 5 Distributed File System
14) Write a note on Trends in Distributed File System.
 Some of the changes that can be expected in the future and some of the implications these
changes may have for file systems are:
1. New Hardware
o Within few years memory may become so cheap that the file system may permanently
reside in memory, and no disks will be needed.
o With an in-core file system, it may be much simpler to store each file contiguously in
memory, rather than breaking it up into blocks.
2. Scalability
o Algorithms that work well for systems with 100 machines may work poorly for systems
with 1000 machines and not at all for systems with 10,000 machines.
o If opening a file requires contacting a single centralized server to record the fact that
the file is open, that server will eventually become a bottleneck as the system grows.
o A general way to deal with this problem is to partition the system into smaller units and
try to make each one relatively independent of the others.
3. Wide Area Networking
o In the future, many LAN-based distributed systems will be interconnected to form
transparent distributed systems covering countries and continents.
o An inherent problem with massive distributed systems is that the network bandwidth
is extremely low.
o Bringing fiber optics into everyone's house will take decades and cost billions.
4. Mobile Users
o A large fraction of the time, the user is off-line, disconnected from the file system.
o Solution is probably going to have to be based on caching.
o While connected, the user downloads to the portable those files expected to be needed
later.
o When reconnect occurs, the files in the cache will have to be merged with those in the
file tree.
o Since disconnect can last for hours or days, the problems of maintaining cache
consistency are much more severe than in online systems.
5. Multimedia
o Text files are rarely more than a few megabytes long, but video files can easily exceed
a gigabyte.
o To handle applications such as video-on-demand, completely different file systems will
be needed.
6. Fault Tolerance
o As distributed systems become more and more widespread, the demand for systems
that essentially never fail will grow.
o File replication, in current distributed systems, will become an essential requirement in
future ones.
o Systems will also have to be designed that manage to function when only partial data
are available, since insisting that all the data be available all the time does not lead to
fault tolerance.

Prof. Dipak Ramoliya (9998771587) | 2160710 – Distributed Operating System 19