The State of the IETF

Download Report

Transcript The State of the IETF

The State of the IETF
Keeping one Internet
Harald Alvestrand, IETF chair
Antalya, May 13, 2001
The IETF in review




What is the Internet?
What is the IETF?
What does the IETF work on?
What challenges do we face?
Thoughts I would like to address

IETF history, structure, and procedure


Relations among standards bodies


Who’s who in the IETF
Who does what and why
Internet directions and concerns
The Internet today

UoSAT-12
The optical internet backbone
Gigabit
to terabit links
Campus
Networks
(LANs)
Internet
in Airlines

Access networks
xDSL,
cable modem, ISDN, asynchronous dial
UNIVERSITY
Internet Engineering Task Force
(IETF)

Historical developer of Internet-related
protocols


Http://www.ietf.org
Consortium of individuals from

Research, Education, Network operators, and
Internet vendors
Fundamental working principle
“We reject kings, presidents, and
voting.
We believe in rough consensus
and running code.
Dr. David C. Clark,
Massachusetts Institute of Technology
”
Fundamental perspective of
enlightened self-interest

There is no one organization or company which
has a corner on intelligence or expertise


Good ideas that help our markets come from
everywhere and anywhere
Therefore, our separate markets grow
interdependently
Example: A better routing algorithm might make network
computers a more acceptable product
How IETF sees work divided
W3C
IEEE
HTML
Voice/ Video
Data
HTTP SNMP
Mail
UDP
RTP
TCP
Internet Protocol
Ethernet ATM Frame Relay
PPP
A variety of physical layers and interfaces
Telephony
Signaling
MPLS
Cellular Radio
ETSI
ITU-T
Applications come from all over
 IETF

Provides
network infrastructure
Tends to use interfaces defined by other bodies
Wants to make sure the whole thing works
IETF: infrastructure applications




SNMP management
SMTP mail
DNS name services
LDAP directory services




Telnet virtual terminal
protocol
FTP file transfer
HTTP web transfer
And more...
IETF vision for the future

Provider view


User view


Internet as interconnected competing service
providers
Internet as universal interconnect
The harmony is not obvious to all
Growth of IP Traffic
Email
 Information
search/access
 Subscription
services/“Push”
 Conferencing/
multimedia
 Video/imaging

“From 2000 on, 80% of Service
Provider Profits Will Be Derived
from IP-Based Services.”
Source: CIMI Corp.
Source: Multiple IXC Projections
Rel. Bit
Volume
250
Traffic Projections for
Voice and Data
Data
(IP)
200
150
Circuit Switched Voice
100
Cross over date
varies with
measuring point
50
1997
1998
1999
2000
2001
Mobile Internet Outlook
Millions
Projected
cellular
subscribers
1,400
1,200
1,000
(Nokia 1999)
More handsets than PCs connected
to the Internet by the end of 2003 !
Projected Web
handsets
800
(Nokia 1999)
600
Projected PCs
connected to
the Internet
400
(Dataquest 10/98)
200
0
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Threat to growth

Balkanization:




Names that can’t be used by all
Formats that can’t be used by all
Networks that can’t be used by all
We need one Internet!
One Protocol: IPv4 and IPv6
Private
Internet
$
IPv6
Internet
IPv4
Internet
V6 and interworking




V6 is deploying (at last)
A plethora of interworking options
A lack of solid experience with usage
Some DNS details being worked on


AAAA vs A6, bitstring labels vs nibbles
Go Build Networks!
One spaghetti: Layer 2 1/2


MPLS, L2TP, ATM, All-over-All
Sub-IP Temporary Area
TE-WG
PPVPN
Others
CCAMP
Control
Measure
MPLS
IPO
ATM
FR
Others
The Sub-IP Temporary Area





7 Working Groups
Shares AD with other areas (Bradner,
Wijnen)
Structure how IP runs over lower layer
media
Includes IP-in-IP, IP-in-MPLS and so on
Does NOT include MPLS-in-MPLS
One Routing Domain





100.000 routes
Probably greatest
short term challenge
Exponential growth
Real requirements
driving growth
Rethink required
Source: Geoff Huston, TELSTRA
One Domain Name System





I18N challenges are more than technical
Identifiers are not names
Getting names into the DNS is the easy part
Patents are a pain
Bq—aervweor3dfae4rtobnlaruoo.com?
Courtesy of i-dns.net
We Want One Internet

Filled with opportunities


All opportunities come with challenges




Global communication enhances business, trade,
research
IPv6 for more addresses
Internationalization for global reach
Scaling routing to a new level
Ours to be responsible with
Questions?