Part III – IP
DOD
TCP/IP Model:
Layers
|
OSI
Model
|
Protocols
(Port
or protocol numbers)
|
Definition
|
|
Process
|
Application
Presentation
Session
|
Telnet
(23)
FTP
(21)
TFTP
(69)
SMTP
(25)
SNMP
(161)
DNS
(53)
BootP
NFS
DHCP
HTTP
(80)
|
Telephone
Network - terminal emulation
File
Transfer Protocol – file transfer that also allows authentication, directory
browsing
Trivial
File Transfer Protocol – stripped down FTP used to backup and restore
routers’ config
Simple
Mail Transfer Protocol – used to send email. POP3 (110) and IMAP (143)
retrieve mail
Simple
Network Management Protocol – collects valuable network info by polling
devices (UDP)
Bootstrap
Protocol – used in diskless stations that receive network info and OS from
the server
Network
File System – allows different file system to interoperate. Uses UDP.
Dynamic
Host Configuration Protocol – enhanced from BootP, can provide IP, subnet,
domain, gateway, DNS and WINS information. Uses UDP.
Hyper
Text Transfer Protocol – WWW protcol
|
|
Host-to-Host
|
Transport
|
TCP
(6)
UDP
(17)
NBP
|
Transmission
Control Protocol – connection-oriented protocol using windowing as
flow-control mechanism. Segments are numbered and the number of the last
segment received is sent back in the acknowledge message.
User
Datagram Protocol – unreliable connection-less protocol that has less
overhead than TCP.
Port
numbers: used by TCP and UDP. Numbers 0-1023 are well-known port numbers.
Numbers 1024 to 65534 can be used by a transmitting host to initiate the
communication.
Name
Binding Protocol – AppleTalk protocol that matches logical device names to
address.
|
|
Internet
|
Network
|
IP
ICMP
(1)
ARP
RARP
Routing
|
Internet
Protocol – four-byte number used to route packets on the internet.
Connectionless Protocol
Internet
Control Message Protocol – management protocol and message svc provider for
IP. Used in “destination unreachable”, “buffer full”, “hop limit” messages,
and in ping and trace. Implemented by all TCP/IP
hosts.
Address
Resolution Protocol – retrieves a MAC address from an IP address
Reverse
Address Resolution Protocol – retrieves an IP address from a MAC address
All
routing protocols operate at this layer
|
|
Network Access
|
Data-Link
Physical
|
Ethernet,
FastEthernet, Token-Ring, FDDI
|
||
IP
Addressing:
Classes:
(n is the network address portion, and
h is the host address portion)
|
Reserved
Addresses:
|
Class
A: n.h.h.h, n ∈[0,127]
(starts with 0) private: 10/8
Class
B: n.n.h.h, n ∈[128.0,191.255]
(starts with 10) private: 172.16/12
Class
C: n.n.n.h, ∈[192.0.0,223.255.255]
(starts with 110) private: 192.168/16
Class
D: multicast
Class
E: research
|
Network
address of 0s : this network or segment
Network
address of 1s : all networks
Host
address of 0s : this host
Host
address of 1s : all hosts
Address
of 1s : all nodes on current network - flooded broadcast
Address
of 0s : used by Cisco to designate the default route
Address
127.0.0.1 : this node used for loopback tests.
|
IP
Subnetting:
Information
|
Formula
|
Mask
|
Number
|
Mask
|
Number
|
Subnet
address
|
yi=(256-number)*i
|
10000000
|
128
|
11111000
|
248
|
First
host
|
yi+1
|
11000000
|
192
|
11111100
|
252
|
Last
host
|
yi+1-2
|
11100000
|
224
|
11111110
|
254
|
Subnet
broadcast address
|
yi+1-1
|
11110000
|
240
|
11111111
|
255
|
i∈ [1,ns]
Number of subnets: ns=2(hostbits-x)
– 2 , – hostbits is the number of bits reserved for the host in that
class (8 for class C, …)
Number of hosts: nh=2x
–2
where x is the number
of unmasked bits
It is essential to know
how to manipulate subnets to create a given number of hosts or subnets. It is
also essential to be able to calculate the broadcast address of a given host or
network and subnet mask.
No comments:
Post a Comment