ECE 6160: Advanced Computer Networks

1
ECE 6160 Advanced Computer Networks

• Networking Components
• Instructor Dr. Xubin (Ben) He
• Email Hexb_at_tntech.edu
• Tel [phone]

2
Prev

• Introduction
• ISPs
• Delays
• Packet Loss

3
Intra-AS Routing vs Inter AS Routing

• Routing within an Autonomous System (AS)
• Also known as Interior Gateway Protocols (IGP)
• Most common Intra-AS routing protocols
• RIP Routing Information Protocol
• OSPF Open Shortest Path First
• IGRP Interior Gateway Routing Protocol (Cisco
  proprietary)

4
Inter-AS routing in the Internet BGP (Border
Gateway Protocol)
BGP provides for routing among autonomous systems
5
Why different Intra- and Inter-AS routing ?

• Policy
• Inter-AS admin wants control over how its
  traffic routed, who routes through its net.
• Intra-AS single admin, so no policy decisions
  needed
• Scale
• hierarchical routing saves table size, reduced
  update traffic
• Performance
• Intra-AS can focus on performance
• Inter-AS policy may dominate over performance

6
Router Architecture Overview

• Two basic router functions
• run routing algorithms/protocol (RIP, OSPF, BGP)
• switching datagrams from incoming to outgoing link

7
High-Level Router Architecture

• Input Ports
• Physical layer functionality, terminates incoming
  physical link
• Interoperates with the data link layer
• Performs lookup and forwarding functions
• In practice, multiple ports are often gathered
  together in a single line card within a router

8
High Level Router Architecture

• Switching Fabric
• Connects the routers input ports to its output
  ports
• Output Ports
• Stores packets forwarded to it through the
  switching fabric
• Routing Processor
• Executes the routing protocols
• Maintains the routing information and forwarding
  tables
• Performs network management functions within the
  router

9
Input Port Functions

• Also known as Decentralized Switching

Physical layer bit-level reception

• A copy of the forwarding table is stored at each
  input port and updated as needed
• The switching decision can be made locally at
  each input port
• Decentralized switching avoids a forwarding
  bottleneck at a single point within the router

Data link layer e.g., Ethernet
10
Complicating Factors

• Backbone routers must operate at high speeds, so
  they therefore must be capable of performing
  millions of lookups per second.
• Line speed a lookup is performed in less than
  the amount of time needed to receive a packet at
  the input port.
• Example Consider an OC48 link that runs at 2.5
  Gbps. Assuming a packet size of 256 bytes, this
  implies a lookup speed of approximately a million
  lookups per second performed.

11
Switching Fabrics
Move packets from the input ports to the output
ports
12
Switching Via Memory

• First generation routers
• packet copied by systems (single) CPU
• speed limited by memory bandwidth (2 bus
  crossings per datagram)

• Modern routers
• input port processor performs lookup, copies
  into memory
• Cisco Catalyst 8500

13
Switching Via a Bus

• Datagram moved from input port memory to output
  port memory via a shared bus
• Switching speed limited by bus bandwidth
• 1 Gbps bus, Cisco 1900 sufficient speed for
  access and enterprise routers (not regional or
  backbone)

14
Switching Via An Interconnection Network

• Overcomes bus bandwidth limitations
• Some interconnection networks were initially
  developed to connect processors in a single
  multiprocessor
• Advanced design fragments datagram into fixed
  length cells, then switches cells through the
  fabric.
• Cisco 12000 switches Gbps through the
  interconnection network

15
Crossbar switching network

• N input and N output buses
• Expensive when N is large
• If the vertical bus is being used to transfer a
  packet from another input port to the same output
  port, the arriving packet is blocked and must be
  queued at the inpit port

16
Multistage Interconnection Network
17
Omega Network
18
Omega Network Features
Omega netowrk is a blocking network. When routes
to different destinations share a link, message
may be blocked by another.
19
Blocking in Omega Networks
Try Omega by youself! (Thanks,Rick)
20
Output Ports

• Transmits the datagrams that have been stored in
  the output ports memory and transports them over
  the outgoing link
• Buffering is required when datagrams arrive from
  fabric faster than the transmission rate of the
  output port
• Scheduling discipline is used to choose among
  queued datagrams for transmission onto network

21
Queuing at the Output Port

• Buffering occurs when arrival rate via the
  switching fabric exceeds output line speed
• Consequently, a delay due to queuing occurs and
  there is potential packet loss due to output port
  buffer overflow

22
Queuing at Input Port

• Switching fabric slower than input ports combined
  means that queueing may occur at input ports
• Head-of-the-Line (HOL) blocking queued datagram
  at front of queue prevents others in queue from
  moving forward
• Consequently, queuing delay and packet loss due
  to input buffer overflow

23
IPv6
24
Addressing Scheme
IPv6 will have 128 bits for the IP address. This
is enough to allow every grain of sand its own IP
address!
25
Addressing Scheme
26
IPv6

• Additional motivation
• header format helps speed processing/forwarding
• new anycast address route to best of several
  replicated servers
• IPv6 datagram format
• fixed-length 40 byte header
• no fragmentation allowed
• ICMPv6 new version of ICMP
• additional message types, e.g. Packet Too Big
• multicast group management functions

27
Header IPv4 vs IPv6
28
IPv6 Header

• A closer look at some of the fields
• Priority identify priority among datagrams in
  flow
• Flow Label identify datagrams in same flow.
• (concept offlow not well
  defined).
• Next header identify upper layer protocol for
  data
• Traffic Class Similar idea to the type of
  service field in IPv4
• Checksum Does not exist in IPv6! It was removed
  entirely to reduce processing time at each
  hop
• Options allowed, but outside of header,
  indicated by Next Header field

29
Transition From IPv4 To IPv6

• Not all routers can be upgraded simultaneously
• no flag days
• How will the network operate with mixed IPv4 and
  IPv6 routers?
• Two proposed approaches
• Dual Stack some routers with dual stack (v6, v4)
  can translate between formats
• Tunneling IPv6 carried as payload in IPv4
  datagram among IPv4 routers

30
Dual Stack Approach
IPv6 nodes have full IPv4 capabilities as well.
When operating with an IPv4 node, the IPv6 node
uses v4 datagrams. The node will be able to
determine the capabilities of the node it is
communicating with by looking at the address
returned by the DNS.
IPv6
IPv6
IPv6
IPv6
IPv4
IPv4
SrcA Dest F data
SrcA Dest F data
A-to-B IPv6
B-to-C IPv4
E-to-F IPv6
D-to-E IPv4
31
Tunneling
tunnel
Logical view
IPv6
IPv6
IPv6
IPv6
Physical view
IPv6
IPv6
IPv6
IPv6
IPv4
IPv4
SrcB Dest E
SrcB Dest E
A-to-B IPv6
E-to-F IPv6
B-to-C IPv6 inside IPv4
D-to-E IPv6 inside IPv4
32
Summary

• Routing
• Router Architecture
• IPV6
• More about IPV6 by Christoph Litz et al.
• Next
• More about Omega Network