Chapter 9 Network Organization Concepts

1
Chapter 9Network OrganizationConcepts

• Understanding Operating Systems, Fourth Edition

2
Objectives

• You will be able to describe
• Several different network topologiesincluding
  the star, ring, bus, tree, and hybrid
• Three types of networks LAN, MAN, and WAN
• The difference between circuit switching and
  packet switching
• Conflict resolution procedures that allow a
  network to share common transmission hardware and
  software effectively

3
Objectives (continued)

• You will be able to describe
• The two transport protocol models (OSI and
  TCP/IP) and how the layers of each one compare

4
Basic Terminology

• Network Collection of loosely coupled processors
  interconnected by communication links using
  cables, wireless technology, or both
• Goal To provide a convenient way to share
  resources (hardware and software) while
  controlling users access to them
• General configurations for OS for networks
• Network operating system (NOS)
• Distributed operating system (D/OS)

5
Basic Terminology (continued)

• Network operating system (NOS) Networking
  capability added to single-user operating system
• Users aware of specific computers and resources
  in the network
• Access via logon to remote host or by data
  transfer from remote host

6
Basic Terminology (continued)

• Distributed operating system (D/OS) Users can
  access remote resources as if local resources
• Good control for distributed computing systems
• Allows resources to be accessed in a unified way
• Represents total view across multiple computer
  systems for controlling and managing resources
  without local dependencies
• Management is a cooperative process
• Comprised of four managers with a wider scope

7
Basic Terminology (continued)

• D/OS must provide the following components
• Process or object management
• Memory management
• File management
• Device management
• Network management

8
Basic Terminology (continued)
Figure 9.1 Networked management system
9
Basic Terminology (continued)

• Advantages of D/OS over traditional systems
• Easy and reliable resource sharing
• Faster computation
• Adequate load balancing
• Good reliability
• Dependable electronic communications among the
  networks users

10
Basic Terminology (continued)

• In distributed system each processor classifies
  other processors and resources as remote and
  considers its own resources local
• Site Indicates a specific location in a network
  with one or more computers
• Host Specific computer system found at a site
  whose services and resources can be used from
  remote locations
• Node Refers to the name assigned to a computer
  system connected to network to identify it to
  other computers in network

11
Basic Terminology (continued)
Figure 9.2 Clients request data or services from
the host server and wait for the response. If the
client host has resources needed by the server
host, the roles can be reversed
12
Network Topologies

• Sites in any networked system can be physically
  or logically connected in a variety of topologies
• Common topologies star, ring, bus, tree, hybrid
• In each topology there are tradeoffs between
• Need for fast communication among all sites
• Tolerance of failure at a site or communication
  link
• Cost of long communication lines
• Difficulty of connecting one site to a large
  number of other sites

13
Network Topologies (continued)

• Four basic criteria
• Basic cost Expense required to link the various
  sites in the system
• Communications cost Time required to send a
  message from one site to another
• Reliability Assurance that many sites can still
  communicate with each other if a link or site
  fails
• Users environment Critical parameters that
  network must meet to be a successful business
  investment

14
Star

• All transmitted data must pass through a central
  controller when going from a sender to a receiver
• Advantages
• Permits easy routing
• Easy access control to the network
• Challenges
• Central site must be extremely reliable and able
  to handle all network traffic, no matter how heavy

15
Star (continued)
Figure 9.3 Star topology
16
Ring

• All sites are connected in a closed loop with the
  first site connected to the last
• Network can be connected to other networks via a
  bridge (same protocols) or gateway (different
  protocols)
• Data is transmitted in packets with source and
  destination address fields
• Each packet is passed from node to node in one
  direction only
• Every node must be functional, or failed node
  needs to be bypassed for proper operation

17
Ring (continued)
Figure 9.4 Ring topology
18
Ring (continued)
Figure 9.5 Double loop computer network using a
ring topology
19
Ring (continued)
Figure 9.6 Multiple rings bridged together
20
Bus

• All sites connected to a single communication
  line
• Messages from any site circulate in both
  directions
• Only one site can successfully send messages at
  one time
• Needs control mechanism to prevent collision
• Data may pass directly from one device to
  another, or it may be routed to an end point
  controller at the end of the line

Figure 9.7 Bus Topology
21
Tree

• Tree A collection of busses connected by a
  branching cable with no closed loops
• Allows users to create networks using bridges
• Message from any site can be received by all
  other sites, until it reaches an end point
• End point controller absorbs a message if it
  reaches end point controller without being
  accepted by a host
• Advantage Message traffic can still flow through
  the network even if a single node fails

22
Tree ( continued)
Figure 9.8 Tree Topology
23
Hybrid
Selects among the strong points of each topology
and combines them to meet that systems
communications requirements most effectively
Figure 9.9 Hybrid topology combining a star and
a ring using a bridge
24
Hybrid (continued)
Figure 9.10 Hybrid topology combining a star and
a bus
25
Network Types

• Grouping of networks according to physical
  distances they cover
• Network types
• Local area networks (LAN)
• Metropolitan area networks (MAN)
• Wide area networks (WAN)

26
Local Area Network

• A configuration found within a single office
  building, campus, or similarly enclosed
  environment
• Owned, used, and operated by single organization
• Allows computers to communicate directly through
  a common communication line
• Communications arent limited to well-defined
  local area only
• LAN can be a component of larger communication
  network
• Provides easy access to outside through bridge or
  gateway

27
Local Area Network (continued)

• Bridge Connects two or more geographically
  distant LANs with same protocols
• e.g., simple bridge used to connect 2 Ethernet
  LANs
• Gateway Connects two or more LANs or systems
  that use different protocols
• Translates one networks protocol into another,
  resolving hardware and software incompatibilities
• e.g., SNA gateway can connect microcomputer
  network to mainframe host

28
Local Area Network (continued)

• Data rates in LAN vary from 100 Mbps to more than
  40 Gbps
• Close physical proximity allows very high-speed
  transmission
• Star, ring, bus, tree, and hybrid are normally
  used to construct local area networks
• Transmission medium used may vary from one
  topology to another
• Factors determining transmission medium include
  cost, data rate, reliability, number of devices
  that can be supported, distance between units etc.

29
Metropolitan Area Network

• Configuration spanning an area larger than a LAN
• Ranging from several blocks of buildings to an
  entire city but not exceeding a circumference of
  100 km
• Owned and operated by a single organization
• Usually used by many individuals organizations
• May be owned and operated as public utilities
  providing means for internetworking several LANs
• MAN high-speed network often configured as a
  logical ring

30
Wide Area Network

• A configuration that interconnects communication
  facilities in different parts of a country or the
  world, or that is operated as part of public
  utility
• Uses communications lines of common carriers
  (e.g., telephone companies)
• Uses broad range of communication media (e.g.,
  satellite, microwaves)
• WANs are generally slower than LANs
• Examples ARPAnet (first WAN), Internet (most
  widely recognized WAN)

31
Wireless Local Area Network

• LAN that uses wireless technology to connect
  computers or workstations located within the
  range of the network
• WLAN typically poses security vulnerabilities

WiMax (802.16) would enable wireless broadband
connections over much greater ranges (up to 10
miles)
Table 9.1 IEEE standards for wireless networks
32
Wireless Local Area Network (continued)
Figure 9.11 Wireless Local Area Network
33
Software Design Issues

• Software issues that must be addressed by network
  designers
• How do sites use addresses to locate other sites?
• How are messages routed and how are they sent?
• How do processes communicate with each other?
• How are conflicting demands for resources
  resolved?

34
Addressing Conventions

• Addressing protocols are closely related to
  network topology and geographic location of each
  site
• Local name Name by which a unit is known within
  its own system
• Global name Name by which a unit is known
  outside its own system
• Must follow standard name lengths, formats, and
  other global conventions

35
Addressing Conventions (continued)

• Domain Name Service (DNS) protocol
• The DNS is hierarchical
• Domain names are read from right to left
• Rightmost portion is the top-level domain
• Next level is the domain name
• Next is one or more subdomain names
• Leftmost portion is the host

36
Routing Strategies

• Router Internetworking device, primarily
  software driven, which directs traffic
• Between two different types of LANs, or
• Between two network segments with different
  protocol addresses
• Operates at Network Layer
• Role of routers changes as network designs change
• Used extensively for connecting sites to each
  other and to Internet

37
Routing Strategies (continued)

• Router functions include
• Securing information generated in predefined
  areas
• Choosing the fastest route from one point to
  another
• Providing redundant network connections
• Routing protocols must consider following
• Addressing
• Address resolution
• Message format
• Error reporting

38
Routing Strategies (continued)

• Message formats allow the protocol to perform its
  functions, such as
• Finding new nodes on a network
• Testing to determine whether theyre working
• Reporting error conditions
• Exchanging routing information
• Establishing connections, and transmitting data
• Most widely used routing protocols on Internet
• Routing information protocol (RIP)
• Open shortest path first (OSPF)

39
Routing Information Protocol

• Selection of a path based on immediate number of
  nodes, or hops, between source and destination
• Path with smallest number of hops chosen always
• Advantages
• Easy to implement
• Disadvantages
• Does not take into consideration bandwidth, data
  priority, or type of network
• Updating and reissuing of routing table every 30
  seconds
• Tables propagate from one router to another

40
Open Shortest Path First

• Selection of a transmission path only after the
  state of a network has been determined
• Routing update messages sent only when changes in
  routing environment occur
• Reduces number of messages in internetwork
• Reduces size of messages by not sending entire
  routing table
• Disadvantages
• Increased memory usage
• Bandwidth savings offset by higher CPU usage for
  shortest path calculation

41
Connection Models

• Types of switching
• Circuit switching
• Packet switching
• Circuit Switching Communication model in which
  dedicated communication path is established
  between two hosts before data transmission begins
• Example Telephone system
• Disadvantage Delay before signal transfer begins
  while the connection is set up

42
Packet Switching

• A store-and-forward technique in which a message
  is divided into multiple equal-sized units
  (packets) before sending to destination
• At destination, packets are reassembled into
  their original long format
• A header containing pertinent information about
  the packet is attached to each packet before
  transmission
• Advantages
• More flexible and more reliable than circuit
  switching
• Provides greater line efficiency
• Allows users to allocate priorities to their
  messages

43
Packet Switching (continued)
Figure 9.12 Packet switching (a) divide the
data into addressed packets (b) send each packet
toward its destination (c) reassemble the data
at the destination
44
Packet Switching (continued)
Table 9.2 Comparison of circuit and packet
switching
45
Packet Switching (continued)

• Methods of selecting the path
• Datagrams
• Virtual circuits
• Datagrams Destination and sequence number of
  packet added to information, uniquely identifying
  message to which packet belongs
• Each packet handled independently and route is
  selected as each packet is accepted into network
• At destination, all packets of same message are
  reassembled

46
Packet Switching (continued)

• Datagrams (continued)
• Message cant be delivered until all packets are
  accounted for
• Receiving node requests retransmission of lost or
  damaged packets
• Advantages
• Helps diminish congestion by sending incoming
  packets through less heavily used paths
• Provides more reliability, because alternate
  paths may be set up when one node fails

47
Packet Switching (continued)

• Virtual Circuit Complete path from sender to
  receiver established before transmission starts
• All packets belonging to a message use same route
• Any node can have several virtual circuits to any
  other node
• Advantage Routing decision made once for all
  packets belonging to same message speeds up
  transmission
• Disadvantages
• If node fails, all virtual circuits using that
  node become unavailable
• Congestion is difficult to resolve when heavy
  traffic

48
Conflict Resolution

• Some method to control access is necessary to
  facilitate equal and fair access to network
• Access control techniques
• Round robin
• Reservation
• Contention
• Medium access control protocols
• Carrier sense multiple access (CSMA)
• Token passing
• Distributed-queue, dual bus

49
Access Control Techniques

• Round Robin A node is given certain amount of
  time to complete transmission, at end of which
  opportunity is passed to next node
• Efficient when many nodes transmitting over long
  periods
• Substantial overhead when few nodes transmit over
  long periods of time
• Reservation Access time on medium is divided
  into slots and node can reserve future time slots
• Well suited for lengthy and continuous traffic

50
Access Control Techniques (continued)

• Reservation (continued)
• Good for a configuration with several terminals
  connected to host computer through single I/O
  port
• Contention No attempt is made to determine whose
  turn it is to transmit nodes compete for access
  to medium
• Major advantage Easy to implement
• Better for short and intermittent traffic
• Works well under light to moderate traffic
• Performance tends to break down under heavy loads

51
CSMA

• Carrier sense multiple access (CSMA)
  Contention-based protocol that is easy to
  implement
• Carrier sense means that a node will listen to,
  or test, communication medium before transmitting
  any messages
• Prevents a collision with another node thats
  currently transmitting
• Multiple access means that several nodes are
  connected to same communication line as peers, on
  the same level, and with equal privileges

52
CSMA (continued)

• Disadvantages of CSMA
• Collision if two or more nodes transmit at same
  instant
• Probability of collisions increases if nodes are
  farther apart
• CSMA less appealing access protocol for large or
  complex networks
• CSMA/CD CSMA algorithm modified to include
  collision detection, e.g., Ethernet
• Collisions not completely eliminated but reduced
• Reduces wasted transmission capacity

53
CSMA (continued)

• CSMA/CD
• Access method prevents multiple nodes from
  colliding during transmission
• e.g., Implemented in LocalTalk, Apples cabling
  system
• If collisions occur, involve only a small packet,
  not actual data (in case of Apple CSMA/CA)
• Protocol does not guarantee data will reach its
  destination, but ensures that any data thats
  delivered will be error free

54
Token Passing

• Special electronic message (token) is generated
  and passed along from node to node
• Only node with the token allowed to transmit, and
  after it has done so, it must pass token on to
  another node
• Fast access collisions are nonexistent
• Typical topologies
• Bus
• Ring

55
Token Passing (continued)

• Token-bus Token is passed to each node in turn,
  which upon receipt, attaches data to it and sends
  to destination
• Receiving node copies data, adds acknowledgment,
  and returns packet to sending node
• Sending node passes token on to next node in
  logical sequence
• Initial node order determined by cooperative
  decentralized algorithm
• Once network is running, turns determined by
  priority based on node activity

56
Token Passing (continued)

• Token-bus (continued)
• Higher overhead at each node than CSMA/CD
• Nodes may have long waits under certain
  conditions before receiving token
• Token-ring Token moves between the nodes in turn
  and in one direction only
• If a node wants to send a message it must wait
  for the free token to come by
• Receiving node copies the message in the packet
  and sets the copied bit to indicate it was
  successfully received

57
DQDB

• Distributed-queue, dual bus (DQDB) Intended for
  use with a dual-bus configuration, where each bus
  transports data in only one direction
• Transmission on each bus consists of a steady
  stream of fixed-size slots
• Slots generated at end of each bus marked free
  and sent downstream, where theyre marked busy
  and written to by nodes ready to transmit
• Nodes read and copy data from slots, which then
  continue to travel toward end of bus, where they
  dissipate

58
DQDB (continued)
Figure 9.13 DQDB protocol
59
DQDB (continued)

• Advantages of DQDB
• Provides negligible delays under light loads and
  predictable queuing under heavy loads
• Suitable for MANs that manage large file
  transfers
• Able to satisfy the needs of interactive users

60
Transport Protocol Standards

• Models intended to address need for universally
  adopted network architecture
• OSI Reference Model
• TCP/IP

61
OSI Reference Model

• Provides basis for connecting open systems for
  distributed applications processing
• Open means that any two systems that conform to
  reference model and related standards can be
  connected, regardless of vendor
• Similar functions collected together into seven
  logical clusters (layers)
• Possible to redesign a layer without affecting
  the adjacent layers
• Handles data transmission from one terminal or
  application program to another

62
OSI Reference Model (continued)

• OSI Reference Model
• At every layer of the sending unit, a new header
  is attached to the previous packet before its
  passed on to the next lower layer
• At the data link layer, a link trailer (LT) is
  added, completing the frame, which is passed to
  the physical layer for transmission
• Receiving unit removes each header or trailer
  until it delivers the data to the application
  program at Layer 7

63
OSI Reference Model (continued)
Figure 9.14 OSI transport protocol model
64
OSI Reference Model (continued)

• Layer 1The Physical Layer Describes all
  mechanical, electrical, and functional
  specifications for connecting a device to a
  particular network
• e.g., 100Base-T, RS449, and CCITT V.35
• Layer 2The Data Link Layer
• Establishes and controls the physical path of
  communications on one side
• Checks for transmission errors and resolves
  problems on the other side
• Typical data link level protocols are HDLC and
  SDLC

65
OSI Reference Model (continued)

• Layer 3The Network Layer Provides services such
  as addressing and routing that move data through
  network to its destination
• Layer 4The Transport Layer Maintains reliable
  data transmission between end users
• Example Transmission Control Protocol (TCP)
• Layer 5The Session Layer Responsible for
• Providing a user-oriented connection service
• Transferring data over communication lines
• Example TCP/IP

66
OSI Reference Model (continued)

• Layer 6The Presentation Layer Responsible for
  data manipulation functions common to many
  applications, such as formatting, compression,
  and encryption.
• Layer 7The Application Layer Application
  programs, terminals, and computers access the
  network at this layer
• Provides interface to users and responsible for
  formatting user data before passing to lower
  layers

67
TCP/IP Model

• Transmission Control Protocol/Internet Protocol
  (TCP/IP)
• Oldest transport protocol standard and the basis
  for Internet communications
• File-transfer protocol to send large files error
  free
• TCP/IP emphasizes internetworking and providing
  connectionless services
• Organizes a communication system with three main
  components processes, hosts, and networks
• TCP/IP model is arranged into four layers

68
TCP/IP Model (continued)
Figure 9.15 TCP/IP model
69
TCP/IP Model (continued)

• Network Access Layer Protocols at this layer
  provide access to a communication network
• Flow control, error control between hosts,
  security, and priority implementation are
  performed at this layer
• Internet Layer Equivalent to portion of network
  layer of OSI model that performs routing
  functions
• Implemented within gateways and hosts
• Example Internet protocol (IP)

70
TCP/IP Model (continued)

• Host-Host Layer Supports mechanisms to transfer
  data between two processes on different host
  computers
• Services include error checking, flow control,
  and an ability to manipulate connection control
  signals
• e.g., Transmission Control Protocol (TCP)
• Process/Application Layer Includes protocols for
  computer-to-computer resource sharing and
  terminal-to-computer remote access
• e.g., FTP, SMTP, and Telnet

71
Summary

• Operating systems for networks necessarily
  include the functions of Memory Manager,
  Processor Manager, Device Manager, and File
  Manager
• Networks operating system must meet the
  reliability requirements of its owners
• Distributed operating systems allows resources to
  be accessed in a unified way
• Sites in any networked system can be physically
  or logically connected to one another in a
  variety of topologies star, ring, bus, tree, and
  hybrid

72
Summary (continued)

• Hybrid topology combines the strong points of
  each topology to meet communications requirements
  most effectively
• Networks are grouped according to physical
  distances they cover LAN, MAN and WAN
• Operating system must detect a failure, change
  routing instructions to avoid that node, and make
  sure every lost message is retransmitted until it
  is successfully received

73
Summary (continued)

• Packet switching provides greater line efficiency
  than circuit switching
• CSMA/CD prevents multiple nodes from colliding
  during transmission
• OSI reference model provides basis for connecting
  open systems for distributed applications
  processing
• TCP/IP is the oldest transport protocol standard
  and the basis for Internet communications