01-intro-history-evolution1

advertisement
Lecture 1
- Practical information
- Internet foundations
- Internet evolution (part 1)
D.Sc. Arto Karila
Helsinki Institute for Information Technology (HIIT)
arto.karila@hiit.fi
10.09.2012
M.Sc. Mark Ain
Helsinki Institute for Information Technology (HIIT)
mark.ain@hiit.fi
T-110.6120 – Special Course in Future Internet Technologies
1
Contents

Practical arrangements


Presentation
Exam
1. Internet foundations
2. Internet evolution (part 1)
10.09.2012
2
Practical arrangements
Welcome to the course!

Staff


Professor:
Assistant:
Arto Karila, D.Sc.
Mark Ain, M.Sc.
arto.karila@hiit.fi
mark.ain@hiit.fi
 We may have guest lecturers for some sessions.

Agenda
1. Internet evolution i.e. problems, architectural and application-based
2.
3.
4.
5.
6.

solutions etc. (10%)
Why the Internet only just works (10%)
Van Jacobson’s NNC: a prominent evolutionary FIA (10%)
Evolutionary and revolutionary future Internet architectures (50%)
LIPSIN demo (10%)
Panel discussion (10%)
Language

English
10.09.2012
3
Practical arrangements
 Credits

4 ECTS (i.e. ~110 hours)
* NOTE: This is an intensive course. You should expect to commit at least
6 hours per week outside of the lectures, and some additional time to
complete your presentation.
 Lecture schedule:


Mon
Tue
16:15 – 18:00
16:15 – 18:00
T5
T5
 Attendance is mandatory (sign attendance sheet before every lecture)
 Prerequisites


Solid understanding of internetworking concepts and technologies
Targeted to graduate and postgraduate students (Bachelor’s minimum)
10.09.2012
4
Practical arrangements

Format + credit breakdown

Lectures (1 ECTS), readings (2 ECTS), presentation (0.5 ECTS), exam (0.5 ECTS)
You may NOT achieve partial credit!

Assessment


PASS/FAIL (you must meet these requirements to pass the course)
 Lecture attendance (mandatory); sign the attendance sheet up front before
every lecture!
 Complete the weekly readings
GRADED COMPONENTS (your final mark will depend EQUALLY on these)
 (33%) Participation in discussions and quality of contributions
 (33%) Presentation (quality of the slides, completeness, quality of ensuing
discussions, your ability to answer questions and lead discussion etc.)
 (33%) Final exam (multiple choice, short answer, essay)
Your final mark will be on the standard numerical 0 – 5 scale.
5
4
3
2
1
0
Excellent
Very good
Good
Fair
Poor
FAIL
90 – 100%
80 – 89%
70 – 79%
60 – 69%
50 – 59%
< 50%
10.09.2012
5
Practical arrangements
 Reading materials
 The weekly readings for the whole course will be
posted on Noppa. You are highly encouraged to
complete them on or ahead of schedule.
 You should read each paper BEFORE the lecture
for which it is assigned.
 EXCEPTION: weeks 39 – 41 (presentations)
10.09.2012
6
Practical arrangements
 Academic honesty

https://into.aalto.fi/display/enregulations/Aalto+University+Code+of+Ac
ademic+Integrity+and+Handling+Violations+Thereof
Dishonest behaviour is defined as practice where the student's
purpose is to give false representation of his/her own or other
student's knowledge and in an attempt to influence the grading of
the course. Examples of dishonest behaviour include cheating on an
exam, copying someone else's work (without providing an adequate
citation), taking an exam for someone else etc.
All cases of academic dishonesty will
be dealt with harshly.
The bottom line: it’s not worth it.
10.09.2012
7
Contents

Practical arrangements


Presentation
Exam
1. Internet foundations
2. Internet evolution (part 1)
10.09.2012
8
Presentation
 You will be required to make a lecture presentation of a
future Internet architecture (FIA) publication of your
choice (7 possibilities, all available on Noppa).
 Presentation = ~60 mins
 Discussion = ~ 30 mins
 Depending on how many students are enrolled, you may
work in groups.
 Requirements…
 Max ~30 slides covering the entire paper but focusing on the
architecture itself, technologies, testing/results, conclusions etc.
 Include approx. 3 discussion topics (e.g. strengths and
weaknesses of the approach, innovative implementations,
socioeconomic considerations, related issues, extrapolations etc.)
 You will give the lecture to your classmates, answer questions,
and lead a discussion during the remaining lecture10.09.2012
time.
9
Presentation
 You will be presenting during weeks 39 – 41
(lectures 5 – 10).
 We will provide a sample presentation in lecture
4 (Van Jacobson’s NNC), one week before the 5th
lecture.
 You have other reading assignments for weeks
39 – 41; you are NOT required to read the FIA
papers, although they WILL be covered on the
final exam; you are encouraged to skim the
papers and take notes during the presentations.
10.09.2012
10
Presentation
We know that this is a demanding task and that two weeks
is not much time to prepare.
 Those who volunteer for an earlier presentation date
and/or a difficult paper will be graded more leniently.
 Your grade will depend mainly on the factors listed in
slide 5; you will NOT be penalized for superficial
reasons.
10.09.2012
11
Presentation
 Papers will be uploaded to Noppa immediately following




the lecture.
E-mail (the names of those who are in your group, along
with) your top 3 choices for a presentation date and
paper to present, to mark.ain@hiit.fi before tomorrow’s
lecture.
You will be assigned a date and paper on a FCFS basis.
Check the lecture schedule on Noppa to see your
assigned lecture and paper.
Send your slides to mark.ain@hiit.fi before your
assigned lecture so they can be posted to Noppa.
10.09.2012
12
Contents

Practical arrangements


Presentation
Exam
1. Internet foundations
2. Internet evolution (part 1)
10.09.2012
13
Exam
 Date and time TBD, approx. 2 hours long
 Format:
 Most likely a combination of multiple choice,
true/false, and short answer (e.g. 2-3 sentences)
questions.
 “open-book” ; you MAY bring any notes you have
taken during the course
 You may NOT bring: readings, books, calculators, or
any other aids that have not specifically been allowed
 You must pass the final exam with a 1 or higher in
order to pass the course!
 IMPORTANT: You must register for the final exam on
Oodi.
10.09.2012
14
In summary…
 Attend all lectures and participate!
 Read all assigned papers; FIA papers optional
(but covered on final exam)
 Presentation
 Take notes (papers, lectures, presentations)
for the final
 Final exam
10.09.2012
15
Contents

Practical arrangements


Presentation
Exam
1. Internet foundations
2. Internet evolution (part 1)
10.09.2012
16
The Internet circa 2006
10.09.2012
17
History of the Internet…

1957:
Advanced Research Projects Agency (ARPA)
was founded after the launch of the Soviet
satellite Sputnik

1968:
ARPA started the development of the
ARPANET

1969:
The first four nodes of the ARPANET were
connected (the first message: ”lo”)

1974:
Robert Kahn and Vinton Cerf defined the basic
Internet architecture (TCP/IP)

1975:
DARPA started the development of Internet
technology

1983:
On 1/1/1983 the whole ARPANET was
converted to TCP/IP
10.09.2012
18
History of the Internet
(cont’d)

1988:
FUNET joined the Internet

1989:
DataNet (by Telecom Finland) was published and BGP-1
defined

1990:
NSFNET was founded

1991:
The first World Wide Web (WWW) client Mosaic was
published at CERN

1993:
CIDR and BGP-4 were adopted

1990’s:
The Internet secured its position as the leading network
architecture

2000:
The number of Internet hosts exceeded 100,000,000
10.09.2012
19
Growth of the Internet
10.09.2012
20
Contents

Practical arrangements


Presentation
Exam
1. Internet foundations
2. Internet evolution (part 1)
10.09.2012
21
Discussion
 When it comes to the Internet’s problems and
solutions, how do we define what is
“evolutionary” and “revolutionary”?
10.09.2012
22
TCP/IP: the first (and only)
revolution
 ARPANET
 Reliable message delivery via 1822 protocol
 Combined addressing and transport via Network
Control Program (NCP)
 By 1982, over 200 nodes  1822/NCP is insufficient…
 Reliability provided by underlying ARPANet
 Open-architecture and federated networking largely
unknown
One node breaks  all application-level
communications break!
10.09.2012
23
TCP/IP: the first (and only)
revolution
 ARPANET was switched entirely from NCP to
TCP/IP on January 1st, 1983.
 Approx. 400 nodes switched overnight!
 Following flag day, ARPANet split:
 Military Network (MILNET)
 Remaining ARPANet for civilian research purposes
10.09.2012
24
Evolutionary approaches
Architectural
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
DNS (~1982)
EGP (precursor to BGP, ~1982)
TCP congestion control (mid-late 1980’s)
CIDR (~1993)
NAT (early 1990’s)
IPv6 (~1995, first RFC 1998)
IPSEC (~1995)
Mobile IP (~1996)
MPLS (~1996)
DiffServ / IntServ (~1998)
HIP (~1999, first RFC 2006)
BGPSec (mid 2000s)
DNSSec (~2004, first deployed at root level ~2010)
10.09.2012
25
Evolutionary approaches
Application-level (next lecture)
Scalable content delivery
1.
1.
2.
3.
Security (confidentiality, anonymity, authentication etc.)
2.
1.
2.
3.
4.
5.
6.
7.
3.
DHTs (~2001)
P2P networks
CDNs (e.g. Akamai)
Asymmetric crypto (e.g. RSA ~1977 or ~1973, Diffie-Hellman ~1976)
PGP (~1991)
SSL/TLS (mid-1990’s, late-1990’s)
PKI (1990’s)
VPNs e.g. PPTP (~1999)
Wireless security e.g. WPA/WPA2/EAP (late 1990’s and beyond)
Tor (mid 2000’s)
Cloud computing
10.09.2012
26
DNS
 Problems: too many hosts, need human-
friendly naming, hosts.txt file lacks scalability
etc.
 Hierarchical and distributed identifier-locator
translation service
 Conceived in ~1982 and deployed in mid1980s
10.09.2012
27
DNS
10.09.2012
28
DNS: quick discussion

What is good about DNS?

What is bad about DNS?

Why is DNS insufficient to support host mobility
and true content-centrism?
10.09.2012
29
EGP (and later, BGP)
 Problems: need for policy routing, autonomous
system segregation, federated networking etc.
 EGP (Exterior Gateway Protocol) is the
predecessor of BGP
 Not to be confused with an exterior gateway protocol
(general term), of which both EGP and BGP are
examples
10.09.2012
30
EGP
 “Distance-vector” reachability protocol
 States
Neighbor acquisition
2. Neighbor reachability
3. Network reachability update
1.
 It was a challenge to prevent loops in EGP;
BGP fixed this
10.09.2012
31
Congestion control (TCP)
 1987: over 10,000 ARPANet hosts  congestion reaching
critical proportions, congestion collapse occurs
frequently
 Congestion collapse: network routing and switching at
full capacity but not completing any useful work
 Problem: congestion
 Solution (Van Jacobson): reduce transmission rates in
response to perceived loss
 Preserve end-to-end principles, relegate functional
changes to endpoints
 Excellent engineering, still in-use today! However,
underlying architectural issue remains almost
wholly unaddressed.
10.09.2012
32
Classless Interdomain
Routing (CIDR)
 Problems: inflexible addressing model, address space
exhaustion
 Before CIDR…
Leading
bits
Class
Size of
network
field
Size of
host
field
Number
of networks
Addresses
per network
16,777,216
Start
address
End address
0.0.0.0
127.255.255.255
Class A
0
8
24
128 (27)
Class B
10
16
16
16,384 (214)
65,536 (216)
128.0.0.0
191.255.255.255
Class C
110
24
8
2,097,152 (221)
256 (28)
192.0.0.0
223.255.255.255
Class D
(multicast)
1110
not
defined
not
defined
not defined
not defined
224.0.0.0
239.255.255.255
Class E
(reserved)
1111
not
defined
not
defined
not defined
not defined
240.0.0.0
255.255.255.255
(224)
 What problems can you observe here?
10.09.2012
33
Classless Interdomain
Routing (CIDR)
 Classfull address allocation and routing was
inefficient!
 After CIDR…
 “Free” subnet mask size
 “Supernetting”
 Reduced size of r0uting tables
 CIDR was a quick fix; it was not meant to
last as long as it has
 CIDR was the last major change to the
Internet’s core architecture
10.09.2012
34
For tomorrow…
 READ:
 D. Clark. 1988. The design philosophy of the
DARPA internet protocols. SIGCOMM Comput.
Commun. Rev. 18, 4 (August 1988), 106-114.
DOI=10.1145/52325.52336
http://doi.acm.org/10.1145/52325.52336
 SUBMIT:
 Presentation group + top 3 choices for date and
paper to mark.ain@hiit.fi (FCFS)
10.09.2012
35
Funded Master’s thesis positions
Area: Future Internet Research
Employer: Helsinki Institute for Information Technology (www.hiit.fi)
FP7 PURSUIT Project
Publish-Subscribe Internetworking Technologies
www.fp7-pursuit.eu
The FP7 PURSUIT project is an EU-funded initiative working to develop a clean-slate redesign of the Internet based on an information-centric publish-subscribe communications paradigm.
Two positions available in the following areas:
Topics
1)
2)
3)
Porting an existing application to the publish-subscribe networking prototype
Information-centric document editing
Design patterns for dynamic and secure service composition in the information space
We are also ready to discuss other topics within Information Centric Networking (ICN).
•
All topics require good programming skills, basic understanding of networking, and motivation to learn new things and work on a clean-slate technology.
•
We are offering students the opportunity to focus on their thesis full-time in a stimulating and international clean-slate project with good tools and adequate
tutoring.
•
The salary will be according to Aalto University policies (currently above 2100 EUR/month).
•
The positions are open immediately and the funding is available until February 2013. The tutoring will continue past February if necessary.
If you are interested, please contact:
D.Sc. Arto Karila
Principal Scientist, HIIT
arto.karila@hiit.fi
D.Sc. Dmitrij Lagutin
Project Manager, HIIT
dmitrij.lagutin@hiit.fi
Open Innovation House (next to TUAS-Talo) - Otaniementie 19-21 - Otaniemi, Espoo
Thank you for your attention!
Questions? Comments?
10.09.2012
37
Download