Sie sind auf Seite 1von 3

Chapter 4

Network Layer
Computer Networks - CS374

Dr. Anwar Ghani A note on the use of these ppt slides:


Lecturer Computer We’re making these slides freely available to all (faculty, students, readers). Computer
Department of Computer Science & Software Engineering Networking: A Top They’re in PowerPoint form so you see the animations; and can add, modify,
and delete slides (including this one) and slide content to suit your needs. Networking: A Top
International Islamic University Islamabad
Down Approach They obviously represent a lot of work on our part. In return for use, we only
ask the following: Down Approach
6th edition ❖ If you use these slides (e.g., in a class) that you mention their source
(after all, we’d like people to use our book!)
6th edition
Jim Kurose, Keith Ross ❖ If you post any slides on a www site, that you note that they are adapted Jim Kurose, Keith Ross
Addison-Wesley from (or perhaps identical to) our slides, and note our copyright of this
material.
Addison-Wesley
March 2012 March 2012
Thanks and enjoy! JFK/KWR

All material copyright 1996-2012 All material copyright 1996-2013


J.F Kurose and K.W. Ross, All Rights Reserved J.F Kurose and K.W. Ross, All Rights Reserved

Network Management 9-1 Network Layer 4-2

1 2

Chapter 4: network layer Chapter 4: outline


chapter goals: 4.1 introduction 4.5 routing algorithms
4.2 virtual circuit and ▪ link state
❖ understand principles behind network layer datagram networks ▪ distance vector
services: ▪ hierarchical routing
4.3 what’s inside a router
▪ network layer service models 4.6 routing in the Internet
4.4 IP: Internet Protocol
▪ forwarding versus routing ▪ datagram format ▪ RIP
▪ how a router works ▪ IPv4 addressing ▪ OSPF
▪ BGP
▪ routing (path selection) ▪ ICMP
▪ broadcast, multicast ▪ IPv6 4.7 broadcast and multicast
routing
❖ instantiation, implementation in the Internet

Network Layer 4-3 Network Layer 4-4

3 4

Router architecture overview Input port functions


two key router functions:
❖ run routing algorithms/protocol (RIP, OSPF, BGP) link
lookup,
forwarding
❖ forwarding datagrams from incoming to outgoing link line layer switch
termination protocol fabric
(receive)
forwarding tables computed, queueing
routing
pushed to input ports routing, management
processor
control plane (software) physical layer:
forwarding data bit-level reception
plane (hardware) data link layer: decentralized switching:
e.g., Ethernet ❖ given datagram dest., lookup output port
high-seed see chapter 5 using forwarding table in input port
switching memory (“match plus action”)
fabric
❖ goal: complete input port processing at
‘line speed’
❖ queuing: if datagrams arrive faster than
router input ports router output ports forwarding rate into switch fabric
Network Layer 4-5 Network Layer 4-6

5 6

1
Switching fabrics Switching via memory
❖ transfer packet from input buffer to appropriate first generation routers:
output buffer ❖ traditional computers with switching under direct control
❖ switching rate: rate at which packets can be of CPU
transfer from inputs to outputs ❖ packet copied to system’s memory

▪ often measured as multiple of input/output line rate ❖ speed limited by memory bandwidth (2 bus crossings per
▪ N inputs: switching rate N times line rate desirable datagram)
❖ three types of switching fabrics
input output
port memory port
memory (e.g., (e.g.,
Ethernet) Ethernet)

system bus
memory bus crossbar

Network Layer 4-7 Network Layer 4-8

7 8

Switching via a bus Switching via interconnection network


❖ overcome bus bandwidth limitations
❖ datagram from input port memory
❖ banyan networks, crossbar, other
to output port memory via a interconnection nets initially
shared bus developed to connect processors in
❖ bus contention: switching speed multiprocessor
limited by bus bandwidth ❖ advanced design: fragmenting
❖ 32 Gbps bus, Cisco 5600: sufficient bus datagram into fixed length cells,
speed for access and enterprise switch cells through the fabric.
crossbar
routers ❖ Cisco 12000: switches 60 Gbps
through the interconnection
network

Network Layer 4-9 Network Layer 4-10

9 10

Output ports This slide in HUGELY important! Output port queueing

datagram
switch buffer link
fabric layer line
protocol termination switch
(send) switch
queueing fabric
fabric

at t, packets more one packet time later


❖ buffering required when datagrams
Datagram arrive
(packets) can be lost from input to output

from fabric faster than the


due to transmission
congestion, lack of buffers
❖ buffering when arrival rate via switch exceeds
rate output line speed
❖ scheduling discipline chooses
Priority among
scheduling – who queued
gets best ❖ queueing (delay) and loss due to output port buffer
datagrams for transmission
performance, network neutrality overflow!
Network Layer 4-11 Network Layer 4-12

11 12

2
Input port queuing
How much buffering?
❖ fabric slower than input ports combined -> queueing may
occur at input queues
❖ RFC 3439 rule of thumb: average buffering equal
▪ queueing delay and loss due to input buffer overflow!
to “typical” RTT (say 250 msec) times link
❖ Head-of-the-Line (HOL) blocking: queued datagram at front
capacity C of queue prevents others in queue from moving forward
▪ e.g., C = 10 Gpbs link: 2.5 Gbit buffer
❖ recent recommendation: with N flows, buffering
equal to
RTT . C switch switch
fabric fabric
N

output port contention: one packet time later:


only one red datagram can be green packet
transferred. experiences HOL
lower red packet is blocked blocking

Network Layer 4-13 Network Layer 4-14

13 14

Questions ?

15

Das könnte Ihnen auch gefallen