A Review of Evolving Network Technology Ethernet & IP With associated infrastructure. J.J. Ekstrom IT 529 Thursday, January 15, 2015

A Review of Evolving Network Technology

Ethernet & IP

With associated infrastructure.

J.J. EkstromIT 529

Thursday, January 15, 2015

Who is winning? Ethernet has won the LAN wars Ethernet is winning the MAN wars

– Utopia, iProvo, Comcast.. Ethernet is contending for part of the WAN… PPOE (Point to

Point over Ethernet) IP has won all best-effort wars wars…

– Most ATM traffic is IP– A large portion of Sonet Traffic is IP

IETF and Vendors making IP transport of choice– Voice over IP – IP Multicast Streaming

Pretty much everything new assumes Ethernet packets with IP in them.

Why?

Simple transports Work faster and cheaper Put the smarts where it can work for more

transports Not as much advantage to smarter

transports

Ethernet Characteristics

Ethernet shared media cable Cable access method (CSMA/CD) Unreliable Packet Delivery Assumes higher layers do most of the work Simple and Relatively fast on whatever

physical transport with any generation of hardware.

Ethernet Shared Media Cable 1

Physics determined the maximum length of the Ethernet cable– signal strength– cable characteristics


All stations (nodes) hook to, and share a single cable


Each station “listens” as it transmits


Each station must transmit a minimum of 64 bytes to “fill” the cable before it stops listening

64 bytes min.


If a 2nd node transmits before the 1st node finishes, the two transmissions collide and they must retransmit

64 bytes min. 64 bytes min.

Ethernet Cable Access Method (CSMA/CD)

CSMA/CD is a media-access method used by Ethernet and 802.3 networks

CSMA/CD stands for Carrier Sense, Multiple Access / Collision Detection

How CSMA/CD Works - 1

A station wishing to transmit first listens for traffic on the cable indicated by a carrier signal (CSMA/CD-Carrier Sense)

Network Cable Carrier Signal


If the carrier signal is detected, the station waits a period of time and tries again

Network Cable Carrier Signal


If NO carrier signal is detected, the station starts transmitting its packet (min. of 64 bytes) and simultaneously listening

Network CableM

IN. O

F 6

4 B

YTE

S


TWO stations can start transmitting at the same time (CSMA/CD - Multiple Access)

Network Cable

MIN

. O

F 6

4 B

YTE

S

MIN

. O

F 6

4 B

YTE

S


If this happens, both stations hear garbage (CSMA/CD - Collision Detection)

Network Cable

MIN

. O

F 6

4 B

YTES

MIN

. O

F 6

4 B

YTES@&*!


When collisons are detected, both stations :– cancel transmissions by sending a jam signal– wait a random amount of time before trying to

transmit again

Network Cable

JAM

SIG

NA

L

JAM

SIG

NA

L

PROBLEM #1

Physics doesn’t allow you to have LAN wires as long as you would like.

SOLUTION #1

Repeater extended wire length, broadcast domain, and collision domain

Repeater

PROBLEM #2

Too many collisions. LAN wouldn’t carry enough traffic.

SOLUTION #2

Bridging segments extends broadcast domain without collisions: Bigger LANs

BRIDGE

PROBLEM #3 Broadcast storms - result from multi-port

bridges “flooding” all ports when packet destination is unknown and a loop exists.

BRIDGE 1

BRIDGE 3 BRIDGE 2

64 bytes min.

Packet returningto original bridge

PROBLEM #3– when the original packet returns to a previous

bridge, new packets are generated and a “storm” is generated.

BRIDGE

BRIDGE BRIDGE

Cycle Repeats

SOLUTION #3

3.1 - 802.1D (spanning tree) installed on bridges.

3.2 - Routers

SOLUTION #3.1

802.1D (Spanning Tree) added to bridges. – Spanning Tree is an algorithm that runs on

bridges to eliminate loops dynamically.

802.1DBRIDGE 1

802.1DBRIDGE 3

802.1DBRIDGE 2

64 bytes min.

802.1D (SpanningTree) determines thatthis link is redundant

and shuts it down

SOLUTION #3.2 Routers - make every segment another

network or subnet by refusing to pass through any packet whose address it does not recognize.

BRIDGE 1

BRIDGE 2

64 bytes min.

RouterBRIDGE 3

SOLUTION #3.2 NOTE:

– in XNS a single broadcast domain is called a “network.”

– in TCP a single broadcast domain is called a “subnet.”

– network personnel often call a collision domain a “segment.”

PROBLEM #4 Topology and failure characteristics -

problems with bus-oriented LANs (i.e., when the wire breaks NONE of the stations can communicate).

SOLUTION #4

Twisted pair LANs.– When any one wire segment fails, the whole

LAN does NOT go down.

Concentrator ConcentratorBridge

Concentrator

PROBLEM #5

Not enough Bandwidth– only 10 MBPS available on each collision

domain

BRIDGE

BRIDGE

BRIDGEConcentrator

Concentrator

Concentrator

SOLUTION #5

Switches (multiport Bridges) - allows more segments (bandwidth) at a lower cost per port.

Concentrator

Concentrator

SWITCH

PROBLEM #6

Controlling User Connectivity– keep groups separate– easily share resources between groups– do adds, moves, and changes without rewiring

SOLUTION #6 VLANs of various forms create isolated

broadcast domains (networks) Connection between Virtual LAN networks

requires a router. People do security in their routers and

firewalls at network boundaries anyway

Problem #7

During roughly the same 20-25 year period Token-Ring LANs, FDDI, ATM, and several other LAN and WAN technologies have been undergoing similar evolutionary tracks as ethernet.

It was not clear that there would be a clear winner. How do you hook them together and protect your

technology investments? Users don’t care how their bits get pushed around,

only that things work.

Solution #7

Internetworking…The real reason IP has won the protocol wars.– Works well on P2P links

– Works well on LANs

– Makes very few demands of participant networks

– “Rough consensus and working code” Motto of the IETF The way to get useful things quickly in a world of confusion…

what works best wins.

Internetworking: Internet, intranets

Outline Best Effort Service ModelGlobal Addressing Scheme

IP Internet

Concatenation of Networks

Protocol Stack

R2

R1

H4

H5

H3H2H1

Network 2 (Ethernet)

Network 1 (Ethernet)

H6

Network 3 (FDDI)

Network 4(point-to-point)

H7 R3 H8

R1

ETH FDDI

IPIP

ETH

TCP R2

FDDI PPP

IP

R3

PPP ETH

IP

H1

IP

ETH

TCP

H8

Service Model Connectionless (datagram-based) Best-effort delivery (unreliable service)

– packets are lost– packets are delivered out of order– duplicate copies of a packet are delivered– packets can be delayed for a long time– (Sound like Ethernet?)

Datagram format Version HLen TOS Length

Ident Flags Offset

TTL Protocol Checksum

SourceAddr

DestinationAddr

Options (variable) Pad(variable)

0 4 8 16 19 31

Data

Problem: Different MTU

All LAN Technologies do not have same maximum packet size.

Network layer has no simple way to determine path

Packets dropped if too big to be forwarded

Solution: Fragmentation and Reassembly

Strategy– fragment when necessary (MTU < Datagram)– try to avoid fragmentation at source host– re-fragmentation is possible – fragments are self-contained datagrams– use CS-PDU (not cells) for ATM– delay reassembly until destination host– do not recover from lost fragments

Example

H1 R1 R2 R3 H8

ETH IP (1400) FDDI IP (1400) PPP IP (512)

PPP IP (376)

PPP IP (512)

ETH IP (512)

ETH IP (376)

ETH IP (512)

Ident = x Offset = 0

Start of header

0

Rest of header

1400 data bytes


Start of header

1

Rest of header

512 data bytes


Start of header

1

Rest of header

512 data bytes


Start of header

0

Rest of header

376 data bytes

Problem: Global Routing

Next hop is always a local decision How do you know which way to send a

packet?

Global Addresses Properties

– globally unique– hierarchical: network + host

Dot Notation– 10.3.2.4– 128.96.33.81– 192.12.69.77

Network Host

7 24

0A:

Network Host

14 16

1 0B:

Network Host

21 8

1 1 0C:

Datagram Forwarding Strategy

– every datagram contains destination’s address– if directly connected to destination network, then forward to host– if not directly connected to destination network, then forward to

some router– forwarding table maps network number into next hop– each host has a default router– each router maintains a forwarding table

Example (R2) Network Number Next Hop 1 R3 2 R1 3 interface 1 4 interface 0

Problem: Network Address binding

Network Layer Address is logical and global MAC addresses are bound to physical network Point-to-Point may have no physical address

Solution: for IPX

Make network address include physical address

16 bit Network number + 48 bit MAC address = 64 bit address

Solution: For IPv4 Map IP addresses into physical addresses

– destination host– next hop router

Techniques– encode physical address in host part of IP address

Assumes fixed host address Doesn’t work with subnets or 48 bit MACs (IP is 32 bits)

– table-based ARP

– table of IP to physical address bindings– broadcast request if IP address not in table– target machine responds with its physical address– table entries are discarded if not refreshed

ARP Details

Request Format– HardwareType: type of physical network (e.g., Ethernet)– ProtocolType: type of higher layer protocol (e.g., IP)– HLEN & PLEN: length of physical and protocol addresses– Operation: request or response – Source/Target-Physical/Protocol addresses

Notes– table entries timeout in about 10 minutes– update table with source when you are the target – update table if already have an entry– do not refresh table entries upon reference

ARP Packet Format

TargetHardwareAddr (bytes 2 – 5)

TargetProtocolAddr (bytes 0 – 3)

SourceProtocolAddr (bytes 2 – 3)

Hardware type = 1 ProtocolType = 0x0800

SourceHardwareAddr (bytes 4 – 5)

TargetHardwareAddr (bytes 0 – 1)

SourceProtocolAddr (bytes 0 – 1)

HLen = 48 PLen = 32 Operation

SourceHardwareAddr (bytes 0 – 3)

0 8 16 31

Solution: IPv6

Make Network Address 128 bits Carry 64 bit IPX addresses Carry 32 bit IP addresses Even carry DEC Net and others But big tables and smart routers!

Internet Control Message Protocol (ICMP)

Echo (ping) Redirect (from router to source host) Destination unreachable (protocol, port, or host) TTL exceeded (so datagrams don’t cycle forever) Checksum failed Reassembly failed Cannot fragment

Problem: Class based

(0)7 bit Class A too few networks, 6 million hosts too many

(10) 15 bit Class B still too few networks, 64,000 hosts still too many.

(110) 23 bit Class C still too few networks 256 hosts too many for many applications.

Address “ownership” companies grow, shrink, die …

Solution: Classless

CIDR – Classless Inter-Domain Routing Block 20 bit network address Class ignored 12 bit host = 4k hosts ISP’s own blocks

Problem: Trust

ISP’s compete for carrier business ISP’s want to give better service to their

own customers Typical routing algorithms require that

routers trust all other routers Rogue routers kill networks

Solution: Different Routing Algorithms

RIP – local routers trust each other OSPF, IGRP, EIGRP– local trust with some

security BGP – Point-to-point manual configuration

Router not obligated to use information.

(How does the Internet ever work?)

Problem: Spanning Tree wastes bandwidth

Blocked links are not used. If they are 10 gig links that is a big deal.

Fail-over times were on the order of 1 minute.

Shutting down the entire spanning tree during recalculation is not acceptable.

Solutions: many small ones

Link aggregation allows redundancy and full use of the bandwidth except during failure.

Rapid Spanning tree allows much faster failover and doesn’t block everything while reconfiguring

Ports connected to end nodes don’t wait at all. (Portfast on cisco)

Problem: Latency in Hierarchy

Datacenters assume that each migration target has similar network performance to other VMs.

Traditional LAN topologies don’t guarantee this.

Solutions: Stir everything (SDN)

Documents

A Review of Evolving Network Technology Ethernet & IP With associated infrastructure. J.J. Ekstrom IT 529 Thursday, January 15, 2015