Internetworking I, Lecture Slide - Computer Science, Slides of Software Engineering

Download Internetworking I, Lecture Slide - Computer Science and more Slides Software Engineering in PDF only on Docsity!

• IP addresses • IP datagram delivery • Internets • Bridged networksTopicsNovember 19, 1998Internetworking I

class26.ppt

CS 213 F’

• 2 –

class26.ppt

The internetworking problem

of networks, orWant to build a single network (an interconnected set

internetwork, or

internet) out of a

– lots of different kinds of networks (Ethernet, FDDI, ATM, wireless,^ •^ heterogeneityChallenges:large collection of separate networks.

• how to unify this hodgepodge?point-to-point)
• how do we provide uniques names for potentially billions of nodes?•^ scale
• how do we find all these nodes? (forwarding and routing)(naming)

Note:

internet refers to a general idea,

Internet refers to

a particular implementation of that idea (IP).

CS 213 F’

• 4 –

class26.ppt

Learning bridges

Problem: Flooding is wasteful A B C X Y Z

bridge

Ethernet A

Ethernet X

port 2 port 1

remembering which hosts are connected to which bridge ports.Optimization: Forward packets only when necessary by learning and

CS 213 F’

• 5 –

class26.ppt

Learning bridges (cont)

1. start with empty hash tableLearning algorithm:

T that maps hosts to ports

2. receive frame from host

src on port p

3. add (src,p) to T
4. delete old entries
5. receive frameForwarding algorithm:

f from host src to host

dst on port p

2. ifT(dst) = n/a then flood

f.

else if T(dst) = p then discard

f

else forward f on port T(dst).

CS 213 F’

• 7 –

class26.ppt

Cycles in bridged networks

to unknown destination1. host writes frame F

B

F1 B1 2. B1 and B2 flood F

B

B B2 F

B F

F2 4. B1 and B2 flood

B

B F

F2 B1 reads F23. B2 reads F1,

B

B F

F1 B2 reads F25. B1 reads F

B

F1 B1 6. B1 and B2 flood F

CS 213 F’

• 8 –

class26.ppt

Spanning tree bridges

B3 A

B

B

B

G

B

B

C

D

E

F

B3 A

B

B

B

G

B

B

C

D

E

F

• (^) “spanning” refers only to networks, not bridgesthat prunes ports. • (^) Tree is constructed dynamically by a distributed “diffusing computation” • (^) Networks are graph nodes, ports are graph edges

CS 213 F’

• 10 –

class26.ppt

Internetworks

Def: An

internetwork (internet for short) is an arbitrary

collection of

physical networks interconnected to

service.provide some sort of host-to-host packet delivery

internetinternet

hosthost

hosthost hosthost

hosthost

CS 213 F’

• 11 –

class26.ppt

Building an internet

XX YY ZZ

network (^2) (ECE)adaptoradaptor adaptoradaptor

adaptoradaptor

AA BB CC

network (^1) (SCS)adaptoradaptor adaptoradaptor

adaptoradaptor

which are at different locations, and possibly built by different vendors.^ We start with two separate, unconnected computer networks (subnets),

Ethernet

ATM

Question: How to present the illusion of one network?

CS 213 F’

• 13 –

class26.ppt

Building an internet (cont)

XX YY ZZ

network (^2) (ECE)adaptoradaptor adaptoradaptor

adaptoradaptor

AA BB C (^) (gateway/rou ter)

C (^) (gateway/rou ter)

network (^1) (SCS)adaptoradaptor adaptoradaptor

adaptoradaptor

adaptoradaptor

128.2. .

IP (^) addresses:routing messages between network1 and network 2 if necessary.on each host and router. IP provides a global name space for the hosts, Finally, we run a software implementation of the Internet Protocol (IP)

128.2. .

128.2.80.

128.2. .

128.2.80.1 128.2.80.2 128.2.80.

class26.ppt

Building an internet (cont)

• CS 213 F’ - 128.2. internetinternet -. - 128.2. -. - 128.2.80.3128.2.80.3 128.2.80.1128.2.80. - 128.2. -.
  • 128.2.80.3128.2. -.
    • 128.2.80. - 128.2. -. - 128.2. -. - 128.2.80.2128.2.80.

CS 213 F’

• 16 –

class26.ppt

IP

• IP service model:

• (^) IP an addressing scheme for all the hosts in the internet– ordering of datagrams not guaranteed (connectionless)– packets can be lost or duplicated (unreliable) – IP tries but doesn’t guarantee that packets will arrive (best effort)packets between two hosts.IP provides best-effort delivery of datagram (connectionless)

• Why would such a limited delivery model be useful?

protocols like TCP and UDP^ •^ provides a basis for building more sophisticated and user-friendlysimple, so it runs on any kind of network

CS 213 F’

• 17 –

class26.ppt

R1 Example internet

R

H H H H4 Network 3 (FDDI) H H H

H

R

(Ethernet)Network 2

(Point-to-point)Network 4

Network 1 (Ethernet)

CS 213 F’

• 19 –

class26.ppt

Ethernet frame^ Encapsulating IP datagrams in Ethernet

headerheaderEthernet frame headerheaderIP datagramIP datagram IP datagram dataIP datagram data

headerheaderIP datagramIP datagram IP datagram dataIP datagram data

Ethernet frameIP datagram

The same idea is used for other types of physical networks

CS 213 F’

• 20 –

class26.ppt

Ver Hlen

IP packet format

VER IP version

HL Header length (in 32-bit words)

TOS Type of service (unused)

Length (^) Datagram length (max 64K B)

ID Unique datagram identifier

Flags (^) xxM (more fragmented packets)

Offset (^) Fragment offset

TTL Time to Live

Protocol Higher level protocol (e.g., TCP)

TOS

Length

Datagram ID (^) Flags

TTL

Protocol

Checksum

Options (variable)Destination IP addressSource IP address

Offset

Data

(^0) (^4) (^8) (^16) (^19) 31