Metro & CPE Flow Router - Monarch Network Architects

Download Report

Transcript Metro & CPE Flow Router - Monarch Network Architects

Internet Evolution into the Future
Dr. Lawrence Roberts
CEO, Founder, Anagran
1
The Beginning of the Internet
ARPANET became the Internet
• 1965 – MIT- 1st Packet Experiment -Roberts
• 1967 - Roberts to ARPA – Designs ARPANET
• 1969 – ARPANET Starts – 1st Packet Network
• 1971 – ARPANET Grows to 18 nodes
ARPANET 1971
Roberts at MIT Computer
• 1983 – TCP/IP installed on ARPANET – Kahn/Cerf
• 1986 – NSF takes over network - NSFNET
• 1991 – Internet opened to commercial use
2
Internet Early History
“Internet”
100,000
Name first used- RFC 675
Roberts term at ARPA
Kahn term at ARPA
Cerf term at ARPA
Hosts or Traffic in bps/10
10,000
SATNET - Satellite to UK
Aloha-Packet Radio
PacketRadioNET
Spans US
DNS
1,000
TCP/IP Design
100
Ethernet
EMAIL
FTP
NCP
TCP/IP
Hosts
Traffic
10
ICCC Demo
X.25 – Virtual Circuit standard
1
1969
1971
1973
1975 1977
1979
1981
1983
1985
1987
3
Original Internet Design
It was designed for Data
File Transfer and Email main activities
Constrained by high cost of memory
–
–
–
–
–
–
–
Only Packet Destination Examined
No Source Checks
No QoS
No Security
Best Effort Only
Voice Considered
Video not feasible
ARPANET July 1977
Not much change since then
4
Changing Use of Internet
Major changes in Network Use
Voice
Video
Totally moving to packets
– Low loss, low delay required
Totally moving to packets
– Low loss, low delay jitter required
Emergency Services
Broadband Edge
No Preference Priority
Must control Edge Traffic
P2P utilizes TCP unfairness – multiple flows
– Congests network – 5% of users take 80% of capacity
5
Network Change Required
Fairness
– Multi-flow applications (P2P) overload access networks
Network Security
– Need User Authentication and Source Checking
Emergency Services
– Need Secure Preference Priorities
Cost & Power
– Growth constrained to Moore’s law & developed areas
Quality
– Video & Voice require lower jitter and loss
6
Internet Traffic
Grown 1012 since 1970
World Internet Traffic - History
10000
1000
P2P
100
PetaBytes per month
10
1
0.1
0.01
0.001
Normal Traffic
0.0001
0.00001
TCP
0.000001
0.0000001
0.00000001
0.000000001
1970
ARPANET
1975
1980
NSFNET
1985
1990
COMMERICAL
1995
2000
2005
In 1999 P2P applications discovered using multiple flows could give them
more capacity and their traffic moved up to 80% of the network capacity
7
Where will the Internet be in the next decade
% World Population On-Line
Total Traffic PB/month
Traffic per User GB/month
GB/mo/user Developed areas
GB/mo/user Less Dev. areas
2008
22%
3,200
2.2
2.7
0.5
2018
99%
191,000
26
156
3
People in less developed areas will have more capacity than is available in
developed areas today!
Users in developed areas could see 3-10 hours of video per day (HD or SD)
Requires a 60 times increase in capacity (Moore’s Law increase)
8
Fairness - In the beginning
A flow was a file transfer, or a voice call
The voice network had 1 flow per user
– All flows were equal (except for 911)
– Early networking was mainly terminal to
computer
– Again we had 1 flow (each way) per user
– No long term analysis was done on fairness
It was obvious that under congestion:
Users are equal
thus
Equal Capacity per Flow
was the default design
9
Fairness - Equal Capacity per Flow in Unfair
The Internet is still equal capacity per flow under
congestion
P2P uses 10-1000 flows – consumes neighbors capacity
P2P
FTP
Typically 1000 ISP users share a capacity pool
Congestion here forces equal capacity per flow
The result is therefore unfair to users who paid the same
Dow nstream Capacity Usage
12 Mbps Dow nstream , 7 Mbps upstream
Traffic % capacity
100%
Equality
Wasted Capacity
80%
60%
P2P Users - 4 Mbps/user
40%
20%
DPI
No Control
Normal Users - 0.3 Mbps/user
0%
5.0%
4.5%
4.0%
3.5%
3.0% 2.5% 2.0%
% Users P2P
1.5%
1.0%
0.5%
0.0%
10
Fairness - Internet Traffic Recently – 80% P2P
Since 2004, total traffic has increased 64% per year, about Moore’s Law
– P2P has increased 91% per year – Consuming most of the capacity growth
– Normal traffic has only increased 22% per year –Significantly slowdown from past
Since P2P slows other traffic 5:1, users can only do 1/5 as much
This may account for the normal traffic being about 1/5 what it should be
with normal growth
World Internet Traffic - History
3500
PetaBytes per month
3000
2500
2000
1500
P2P
1000
500
0
2000
Normal Traffic
2002
2004
2006
11
2008
Fairness - A New Rule is Needed
Today - Equal capacity per flow
– P2P has taken advantage of this, using 10-1000 flows
– This gives the 5% P2P users 80-95% of the capacity
– P2P does not know when to stop until it sees congestion
Instead we should give Equal capacity for equal pay
– This is simply a revised equality rule – similar users get equal capacity
– This tracks with what we pay
P2P
This is a major worldwide problem
FTP
– P2P is not bad, it can be quite effective
– But, without revised fairness, multi-flow applications will proliferate
– It then becomes an arms race – who can use the most flows
12
Security – User Authentication in the Network
Today – All security is left to the computer
– The network does not even verify the source address
– There is no way to determine who sent Spam or Malware
Goal – Network to secure each connection (flow)
– User and computer ID sent to network to be verified
– Network to verify the source address is correct
Also, allow the receiver to verify the senders ID
– If no legal ID, connection can be refused
This greatly improves Internet security
– Source of Malware can be recorded and controlled
– Computer security becomes much easier
Protocol for this is now proceeding in the ITU
13
Emergency Services – Required on Internet
Today – Telephone system has GETS for emergencies
– However, Internet taking over voice and has no such service
Required – Both priority and user authentication security
– Priority is not available in the Internet – it could be misused
– Secure user authentication required before priority offered
Priority must provide higher rates, not just lower delay
Same problem as Multi-Flow fairness – needs rate priority
– Standard queue based discard for congestion cannot do it
– Flow based rate control is required
Solution requires change at network edge
– Flow rate control and per flow user authentication
14
A New Alternative - Flow Rate Management
Network Equipment now drops random packets
– All traffic suffers delay and jitter if there is any congestion
– Voice & Video do not slow down but still lose packets
– Data flows often lose several packets and stall
Flow Rate Management - a new control alternative
–
–
–
–
Control the rate of each TCP flow individually
Smoothly adjust the TCP rates to fill the desired capacity
Insure congestion does not occur by controlling rates
Provide for user authentication and emergency service priorities
Replacing random drops with rate control:
–
–
–
–
Network Stability is maintained
All traffic moves smoothly without random loss and stalls
Voice & Video flows cleanly with no loss or delay jitter
Unfairness can be eliminated
15
Flow Rate Management – Why now?
Routers, WAN Optimizers, & DPI depend on processing
– They process every packet – there are 14 packets per flow
Memory cost has come down faster than processing cost
Flow Rate Management is memory based (flow table)
– Stream packets without delay, rate control flows
Allows it to support four 10 Gbps trunks in 1 RU
16
Cost and Power – Reduction help Internet Grow
Today – Network equipment is packet based
– Every packet re-examined – Extensive processing
Flow Rate Management processes flows
– Hardware lookup of flow record for incoming packets
– Flow rate measured and, if required, controlled
– Output rates measured, feedback insure no congestion
Result - comparing flow vs. packet processing:
– Power, size and cost 3-5 times lower managing flows
– New capabilities available: fairness, security, priority, quality
These factors allow a step-up in Internet capacity
– Cost (4:1). Fairness (5:1). Congestion (2:1) – Total 40:1
17
Impact - Optical Network
Today – Fiber cost has crossed below Equipment cost
– Fiber capacity cost has been reducing by half every 12 months
– Equipment cost has been reducing b half every 18 months
This limits the Internet capacity growth to Moore’s Law
– Equipment dominates, Income/user fixed, Thus growth limited
Flow Rate Management permits a major cost reduction
– The network edge $/Gbps can be reduced by over 10:1
– The network core can use primarily optical switching
This can let the Internet capacity double each year again
Lower power is also critical for the environment
18