1. Chapter 2
Application Layer
A note on the use of these ppt slides:
We’re making these slides freely available to all (faculty, students, readers).
They’re in PowerPoint form so you can add, modify, and delete slides
(including this one) and slide content to suit your needs. They obviously Computer Networking:
represent a lot of work on our part. In return for use, we only ask the A Top Down Approach,
following:
If you use these slides (e.g., in a class) in substantially unaltered form, 5th edition.
that you mention their source (after all, we’d like people to use our book!) Jim Kurose, Keith Ross
Addison-Wesley, April
If you post any slides in substantially unaltered form on a www site, that
you note that they are adapted from (or perhaps identical to) our slides, and
note our copyright of this material. 2009.
Thanks and enjoy! JFK/KWR
All material copyright 1996-2009
J.F Kurose and K.W. Ross, All Rights Reserved
2: Application Layer 1
2. Chapter 2: Application layer
Ì 2.1 Principles of Ì 2.6 P2P applications
network applications Ì 2.7 Socket programming
Ì 2.2 Web and HTTP with TCP
Ì 2.3 FTP Ì 2.8 Socket programming
Ì 2.4 Electronic Mail with UDP
SMTP, POP3, IMAP
Ì 2.5 DNS
2: Application Layer 2
3. Chapter 2: Application Layer
Our goals: Ì learn about protocols
Ì conceptual, by examining popular
implementation application-level
aspects of network protocols
application protocols HTTP
transport-layer FTP
service models SMTP / POP3 / IMAP
client-server
DNS
paradigm Ì programming network
peer-to-peer applications
socket API
paradigm
(tutorial)
2: Application Layer 3
4. Some network apps
Ì e-mail Ì voice over IP
Ì web Ì real-time video
Ì instant messaging conferencing
Ì remote login Ì grid computing
Ì P2P file sharing Ì
Ì multi-user network Ì
games Ì
Ì streaming stored video
clips
2: Application Layer 4
5. Creating a network app application
transport
network
data link
write programs that physical
run on (different) end
systems
communicate over network
e.g., web server software
communicates with browser
software application
transport
No need to write software network
data link
for network-core devices
application
physical
transport
network
Network-core devices do data link
physical
not run user applications
applications on end systems
allows for rapid app
development, propagation
2: Application Layer 5
7. Client-server architecture
server:
always-on host
permanent IP address
server farms for
scaling
clients:
client/server
communicate with server
may be intermittently
connected
may have dynamic IP
addresses
do not communicate
directly with each other
2: Application Layer 7
8. Pure P2P architecture
Ì no always-on server
Ì arbitrary end systems
directly communicate peer-peer
Ì peers are intermittently
connected and change IP
addresses
Highly scalable but
difficult to manage
2: Application Layer 8
9. Hybrid of client-server and P2P
Skype
voice-over-IP P2P application
centralized server: finding address of remote
party:
client-client connection: direct (not through
server)
Instant messaging
chatting between two users is P2P
centralized service: client presence
detection/location
• user registers its IP address with central
server when it comes online
• user contacts central server to find IP
addresses of buddies
2: Application Layer 9
10. Processes communicating
Process: program running Client process: process
within a host. that initiates
Ì within same host, two communication
processes communicate Server process: process
using inter-process that waits to be
communication (defined contacted
by OS). Ì Note: applications with
Ì processes in different P2P architectures have
hosts communicate by client processes &
exchanging messages server processes
2: Application Layer 10
11. Sockets
host or host or
Ì process sends/receives server server
messages to/from its
socket controlled by
app developer
Ì socket analogous to door process process
sending process shoves socket socket
message out door TCP with TCP with
buffers, Internet buffers,
sending process relies on variables variables
transport infrastructure
on other side of door which
controlled
brings message to socket by OS
at receiving process
Ì API: (1) choice of transport protocol; (2) ability to fix
a few parameters (more on this in tutorial)
2: Application Layer 11
12. Addressing processes
Ì to receive messages,
process must have
identifier
Ì host device has unique
32-bit IP address
Ì Q: does IP address of
host suffice for
identifying the process?
2: Application Layer 12
13. Addressing processes
Ì to receive messages, Ì identifier includes both
process must have IP address and port
identifier numbers associated with
Ì host device has unique process on host.
32-bit IP address Ì Example port numbers:
Ì Q: does IP address of HTTP server: 80
host on which process Mail server: 25
runs suffice for Ì to send HTTP message
identifying the to gaia.cs.umass.edu web
process? server:
A: No, many IP address: 128.119.245.12
processes can be Port number: 80
running on same host Ì more shortly…
2: Application Layer 13
14. App-layer protocol defines
Ì Types of messages Public-domain protocols:
exchanged, Ì defined in RFCs
e.g., request, response Ì allows for
Ì Message syntax: interoperability
what fields in messages &
Ì e.g., HTTP, SMTP
how fields are delineated
Ì Message semantics
Proprietary protocols:
meaning of information in
Ì e.g., Skype
fields
Ì Rules for when and how
processes send &
respond to messages
2: Application Layer 14
15. What transport service does an app need?
Data loss Throughput
Ì some apps (e.g., audio) can Ì some apps (e.g.,
tolerate some loss multimedia) require
Ì other apps (e.g., file minimum amount of
transfer, telnet) require throughput to be
100% reliable data “effective”
transfer Ì other apps (“elastic apps”)
Timing make use of whatever
Ì some apps (e.g., throughput they get
Internet telephony,
interactive games) Security
require low delay to be Ì Encryption, data integrity,
“effective” …
2: Application Layer 15
16. Transport service requirements of common apps
Application Data loss Throughput Time Sensitive
file transfer no loss elastic no
e-mail no loss elastic no
Web documents no loss elastic no
real-time audio/video loss-tolerant audio: 5kbps-1Mbps yes, 100’s msec
video:10kbps-5Mbps
stored audio/video loss-tolerant same as above yes, few secs
interactive games loss-tolerant few kbps up yes, 100’s msec
instant messaging no loss elastic yes and no
2: Application Layer 16
17. Internet transport protocols services
TCP service: UDP service:
Ì connection-oriented: setup Ì unreliable data transfer
required between client and between sending and
server processes receiving process
Ì reliable transport between Ì does not provide:
sending and receiving process connection setup,
Ì flow control: sender won’t reliability, flow control,
overwhelm receiver congestion control, timing,
throughput guarantee, or
Ì congestion control: throttle
security
sender when network
overloaded
Ì does not provide: timing, Q: why bother? Why is
minimum throughput there a UDP?
guarantees, security
2: Application Layer 17
18. Internet apps: application, transport protocols
Application Underlying
Application layer protocol transport protocol
e-mail SMTP [RFC 2821] TCP
remote terminal access Telnet [RFC 854] TCP
Web HTTP [RFC 2616] TCP
file transfer FTP [RFC 959] TCP
streaming multimedia HTTP (eg Youtube), TCP or UDP
RTP [RFC 1889]
Internet telephony SIP, RTP, proprietary
(e.g., Skype) typically UDP
2: Application Layer 18
19. Web and HTTP
First some jargon
Ì Web page consists of objects
Ì Object can be HTML file, JPEG image, Java
applet, audio file,…
Ì Web page consists of base HTML-file which
includes several referenced objects
Ì Each object is addressable by a URL
Ì Example URL:
www.someschool.edu/someDept/pic.gif
host name path name
2: Application Layer 19
20. HTTP overview
HTTP: hypertext
transfer protocol HT
TP
re q
Ì Web’s application layer PC running HT ues
TP t
protocol Explorer re s
pon
se
Ì client/server model
client: browser that
st
requests, receives, re que
TP se Server
T p on
“displays” Web objects H res running
T P Apache Web
server: Web server HT
server
sends objects in
response to requests
Mac running
Navigator
2: Application Layer 20
21. HTTP overview (continued)
Uses TCP: HTTP is “stateless”
Ì client initiates TCP Ì server maintains no
connection (creates socket) information about
to server, port 80 past client requests
Ì server accepts TCP
connection from client aside
Protocols that maintain
Ì HTTP messages (application- “state” are complex!
layer protocol messages) Ì past history (state) must
exchanged between browser be maintained
(HTTP client) and Web
Ì if server/client crashes,
server (HTTP server)
their views of “state” may
Ì TCP connection closed
be inconsistent, must be
reconciled
2: Application Layer 21
22. HTTP connections
Nonpersistent HTTP Persistent HTTP
Ì At most one object is Ì Multiple objects can
sent over a TCP be sent over single
connection. TCP connection
between client and
server.
2: Application Layer 22
23. Nonpersistent HTTP
(contains text,
Suppose user enters URL references to 10
www.someSchool.edu/someDepartment/home.index jpeg images)
1a. HTTP client initiates TCP
connection to HTTP server
(process) at
1b. HTTP server at host
www.someSchool.edu waiting
www.someSchool.edu on port 80
for TCP connection at port 80.
“accepts” connection,
notifying client
2. HTTP client sends HTTP
request message (containing
URL) into TCP connection 3. HTTP server receives request
socket. Message indicates message, forms response
that client wants object message containing requested
someDepartment/home.index object, and sends message
into its socket
time
2: Application Layer 23
24. Nonpersistent HTTP (cont.)
4. HTTP server closes TCP
connection.
5. HTTP client receives response
message containing html file,
displays html. Parsing html
file, finds 10 referenced jpeg
objects
time 6. Steps 1-5 repeated for each
of 10 jpeg objects
2: Application Layer 24
25. Non-Persistent HTTP: Response time
Definition of RTT: time for
a small packet to travel
from client to server
and back. initiate TCP
connection
Response time: RTT
Ì one RTT to initiate TCP request
file
connection time to
RTT
transmit
Ì one RTT for HTTP
file
request and first few file
received
bytes of HTTP response
to return time time
Ì file transmission time
total = 2RTT+transmit
time 2: Application Layer 25
26. Persistent HTTP
Nonpersistent HTTP issues: Persistent HTTP
Ì requires 2 RTTs per object Ì server leaves connection
Ì OS overhead for each TCP open after sending
connection response
Ì browsers often open parallel Ì subsequent HTTP messages
TCP connections to fetch between same
referenced objects client/server sent over
open connection
Ì client sends requests as
soon as it encounters a
referenced object
Ì as little as one RTT for all
the referenced objects
2: Application Layer 26
27. HTTP request message
Ì two types of HTTP messages: request, response
Ì HTTP request message:
ASCII (human-readable format)
request line
(GET, POST, GET /somedir/page.html HTTP/1.1
HEAD commands) Host: www.someschool.edu
User-agent: Mozilla/4.0
header Connection: close
lines Accept-language:fr
Carriage return,
(extra carriage return, line feed)
line feed
indicates end
of message
2: Application Layer 27
29. Uploading form input
Post method:
Ì Web page often
includes form input URL method:
Ì Input is uploaded to Ì Uses GET method
server in entity body Ì Input is uploaded in
URL field of request
line:
www.somesite.com/animalsearch?monkeys&banana
2: Application Layer 29
30. Method types
HTTP/1.0 HTTP/1.1
Ì GET Ì GET, POST, HEAD
Ì POST Ì PUT
Ì HEAD uploads file in entity
body to path specified
asks server to leave
in URL field
requested object out of
response Ì DELETE
deletes file specified in
the URL field
2: Application Layer 30
31. HTTP response message
status line
(protocol
status code HTTP/1.1 200 OK
status phrase) Connection close
Date: Thu, 06 Aug 1998 12:00:15 GMT
header Server: Apache/1.3.0 (Unix)
lines Last-Modified: Mon, 22 Jun 1998 …...
Content-Length: 6821
Content-Type: text/html
data, e.g., data data data data data ...
requested
HTML file
2: Application Layer 31
32. HTTP response status codes
In first line in server->client response message.
A few sample codes:
200 OK
request succeeded, requested object later in this message
301 Moved Permanently
requested object moved, new location specified later in
this message (Location:)
400 Bad Request
request message not understood by server
404 Not Found
requested document not found on this server
505 HTTP Version Not Supported
2: Application Layer 32
33. HTTP Server
How to scale to large number of users?
-Fork
--create a copy of itself which handles each
client request
-Threads
--create a thread to handle each request
34. HTTP Server
Forked server operation
--A single listen socket for client requests
--A connected socket to each request
--A forked process for each connected
socket
---child process closes listening socket
(Why?)
--Server goes back to listening
--Forked process exits after connection is
processed.
35. HTTP Server
Threaded server
--Listening is similar
--connection socket created
--A thread is given the connection socket
---listening socket is left alone by thread
--Thread closes connected socket after
request
--Server keeps listening for new requests
36. HTTP Client
How to reduce fetch time?
-Pipelining
--send multiple requests without waiting for
responses –persistent connections
--Pipelined requests are expected to be served
“in-order” to client
--Use headers to separate responses
-Multiple requests sent via
--Threads or Forked processes
37. Trying out HTTP (client side) for yourself
1. Telnet to your favorite Web server:
telnet cis.poly.edu 80 Opens TCP connection to port 80
(default HTTP server port) at cis.poly.edu.
Anything typed in sent
to port 80 at cis.poly.edu
2. Type in a GET HTTP request:
GET /~ross/ HTTP/1.1 By typing this in (hit carriage
Host: cis.poly.edu return twice), you send
this minimal (but complete)
GET request to HTTP server
3. Look at response message sent by HTTP server!
2: Application Layer 37
38. User-server state: cookies
Example:
Many major Web sites
use cookies Ì Susan always access
Four components: Internet always from PC
1) cookie header line of Ì visits specific e-
HTTP response message commerce site for first
2) cookie header line in time
HTTP request message
3) cookie file kept on Ì when initial HTTP
user’s host, managed by requests arrives at site,
user’s browser site creates:
4) back-end database at
Web site unique ID
entry in backend
database for ID
2: Application Layer 38
40. Cookies (continued)
aside
What cookies can bring: Cookies and privacy:
Ì authorization Ì cookies permit sites to
Ì shopping carts learn a lot about you
Ì you may supply name
Ì recommendations
and e-mail to sites
Ì user session state
(Web e-mail)
How to keep “state”:
Ì protocol endpoints: maintain state
at sender/receiver over multiple
transactions
Ì cookies: http messages carry state
2: Application Layer 40
41. Web caches (proxy server)
Goal: satisfy client request without involving origin server
Ì user sets browser: origin
server
Web accesses via
cache HT Proxy
TP st
Ì browser sends all req server re que
HT ues TP e
client TP t T ons
HTTP requests to res H e sp
pon
se TPr
HT
cache ues
t
eq se
object in cache: cache T Pr on
HT p
returns object P re s
H TT
else cache requests
object from origin client
origin
server, then returns server
object to client
2: Application Layer 41
42. More about Web caching
Ì cache acts as both Why Web caching?
client and server Ì reduce response time
Ì typically cache is for client request
installed by ISP Ì reduce traffic on an
(university, company, institution’s access
residential ISP) link.
Ì Internet dense with
caches: enables “poor”
content providers to
effectively deliver
content (but so does
P2P file sharing)
2: Application Layer 42
43. Caching example
origin
Assumptions servers
Ì average object size = 100,000
public
bits Internet
Ì avg. request rate from
institution’s browsers to
origin servers = 15/sec
1.5 Mbps
Ì delay from institutional router access link
to any origin server and back
institutional
to router = 2 sec network
10 Mbps LAN
Consequences
Ì utilization on LAN = 15%
Ì utilization on access link = 100%
institutional
Ì total delay = Internet delay + cache
access delay + LAN delay
= 2 sec + minutes + milliseconds
2: Application Layer 43
44. Caching example (cont)
origin
possible solution servers
Ì increase bandwidth of access
public
link to, say, 10 Mbps Internet
consequence
Ì utilization on LAN = 15%
Ì utilization on access link = 15% 10 Mbps
Ì Total delay = Internet delay + access link
access delay + LAN delay institutional
= 2 sec + msecs + msecs network
10 Mbps LAN
Ì often a costly upgrade
institutional
cache
2: Application Layer 44
45. Caching example (cont)
origin
possible solution: install servers
cache public
Ì suppose hit rate is 0.4 Internet
consequence
Ì 40% requests will be
satisfied almost immediately
1.5 Mbps
Ì 60% requests satisfied by
access link
origin server
Ì utilization of access link institutional
reduced to 60%, resulting in network
10 Mbps LAN
negligible delays (say 10
msec)
Ì total avg delay = Internet
delay + access delay + LAN institutional
delay = .6*(2.01) secs + .
cache
4*milliseconds < 1.4 secs
2: Application Layer 45
46. Conditional GET
Ì Goal: don’t send object if cache server
cache has up-to-date cached HTTP request msg
version If-modified-since:
object
Ì cache: specify date of <date>
not
cached copy in HTTP request modified
HTTP response
If-modified-since: HTTP/1.0
<date> 304 Not Modified
Ì server: response contains no
object if cached copy is up-
HTTP request msg
to-date: If-modified-since:
HTTP/1.0 304 Not <date> object
Modified modified
HTTP response
HTTP/1.0 200 OK
<data>
2: Application Layer 46
47. FTP: the file transfer protocol
FTP file transfer
FTP FTP
user client server
interface
user
at host remote file
local file system
system
Ì transfer file to/from remote host
Ì client/server model
client: side that initiates transfer (either to/from
remote)
server: remote host
Ì ftp: RFC 959
Ì ftp server: port 21
2: Application Layer 47
48. FTP: separate control, data connections
TCP control connection
Ì FTP client contacts FTP server port 21
at port 21, TCP is transport
protocol TCP data connection
Ì client authorized over control FTP port 20 FTP
connection client server
Ì client browses remote Ì server opens another TCP
directory by sending commands
data connection to transfer
over control connection.
another file.
Ì when server receives file Ì control connection: “out of
transfer command, server
band”
opens 2nd TCP connection (for
Ì FTP server maintains “state”:
file) to client
current directory, earlier
Ì after transferring one file,
authentication
server closes data connection.
2: Application Layer 48
49. FTP commands, responses
Sample commands: Sample return codes
Ì sent as ASCII text over Ì status code and phrase (as
control channel in HTTP)
Ì USER username Ì 331 Username OK,
Ì PASS password password required
Ì LIST return list of file in Ì 125 data connection
current directory already open;
transfer starting
Ì RETR filename retrieves
Ì 425 Can’t open data
(gets) file connection
Ì STOR filename stores Ì 452 Error writing
(puts) file onto remote file
host
2: Application Layer 49
50. Electronic Mail outgoing
message queue
user mailbox
user
Three major components: agent
Ì user agents mail
user
Ì mail servers server
agent
Ì simple mail transfer SMTP mail
protocol: SMTP server user
SMTP agent
User Agent
Ì a.k.a. “mail reader” SMTP
mail user
Ì composing, editing, reading agent
server
mail messages
Ì e.g., Eudora, Outlook, elm, user
Mozilla Thunderbird agent
user
Ì outgoing, incoming messages agent
stored on server
2: Application Layer 50
51. Electronic Mail: mail servers
user
Mail Servers agent
Ì mailbox contains incoming
mail
user
messages for user server
agent
Ì message queue of outgoing
SMTP
(to be sent) mail messages mail
server user
Ì SMTP protocol between mail
servers to send email SMTP agent
messages SMTP
client: sending mail mail user
agent
server server
“server”: receiving mail
user
server agent
user
agent
2: Application Layer 51
52. Electronic Mail: SMTP [RFC 2821]
Ì uses TCP to reliably transfer email message from client
to server, port 25
Ì direct transfer: sending server to receiving server
Ì three phases of transfer
handshaking (greeting)
transfer of messages
closure
Ì command/response interaction
commands: ASCII text
response: status code and phrase
Ì messages must be in 7-bit ASCII
2: Application Layer 52
53. Scenario: Alice sends message to Bob
1) Alice uses UA to compose 4) SMTP client sends Alice’s
message and “to” message over the TCP
bob@someschool.edu connection
2) Alice’s UA sends message 5) Bob’s mail server places the
to her mail server; message message in Bob’s mailbox
placed in message queue 6) Bob invokes his user agent
3) Client side of SMTP opens to read message
TCP connection with Bob’s
mail server
1 mail
mail
server user
user server
2 agent
agent 3 6
4 5
2: Application Layer 53
54. Sample SMTP interaction
S: 220 hamburger.edu
C: HELO crepes.fr
S: 250 Hello crepes.fr, pleased to meet you
C: MAIL FROM: <alice@crepes.fr>
S: 250 alice@crepes.fr... Sender ok
C: RCPT TO: <bob@hamburger.edu>
S: 250 bob@hamburger.edu ... Recipient ok
C: DATA
S: 354 Enter mail, end with "." on a line by itself
C: Do you like ketchup?
C: How about pickles?
C: .
S: 250 Message accepted for delivery
C: QUIT
S: 221 hamburger.edu closing connection
2: Application Layer 54
55. Try SMTP interaction for yourself:
Ì telnet servername 25
Ì see 220 reply from server
Ì enter HELO, MAIL FROM, RCPT TO, DATA, QUIT
commands
above lets you send email without using email client
(reader)
2: Application Layer 55
56. SMTP: final words
Ì SMTP uses persistent Comparison with HTTP:
connections
Ì HTTP: pull
Ì SMTP requires message
Ì SMTP: push
(header & body) to be in 7-
bit ASCII Ì both have ASCII
Ì SMTP server uses command/response
CRLF.CRLF to determine interaction, status codes
end of message
Ì HTTP: each object
encapsulated in its own
response msg
Ì SMTP: multiple objects
sent in multipart msg
2: Application Layer 56
57. Mail message format
SMTP: protocol for
exchanging email msgs header
blank
RFC 822: standard for text
line
message format:
Ì header lines, e.g.,
To:
body
From:
Subject:
different from SMTP
commands!
Ì body
the “message”, ASCII
characters only
2: Application Layer 57
58. Mail access protocols
SMTP SMTP access user
user
agent protocol agent
sender’s mail receiver’s mail
server server
Ì SMTP: delivery/storage to receiver’s server
Ì Mail access protocol: retrieval from server
POP: Post Office Protocol [RFC 1939]
• authorization (agent <-->server) and download
IMAP: Internet Mail Access Protocol [RFC 1730]
• more features (more complex)
• manipulation of stored msgs on server
HTTP: gmail, Hotmail, Yahoo! Mail, etc.
2: Application Layer 58
59. POP3 protocol S: +OK POP3 server ready
C: user bob
authorization phase S: +OK
C: pass hungry
Ì client commands: S: +OK user successfully logged on
user: declare username C: list
pass: password S: 1 498
Ì server responses S: 2 912
+OK
S: .
C: retr 1
-ERR S: <message 1 contents>
transaction phase, client: S: .
C: dele 1
Ì list: list message numbers C: retr 2
Ì retr: retrieve message by S: <message 1 contents>
number S: .
C: dele 2
Ì dele: delete
C: quit
Ì quit S: +OK POP3 server signing off
2: Application Layer 59
60. POP3 (more) and IMAP
More about POP3 IMAP
Ì Previous example uses Ì Keep all messages in
“download and delete” one place: the server
mode. Ì Allows user to
Ì Bob cannot re-read e- organize messages in
mail if he changes folders
client Ì IMAP keeps user state
Ì “Download-and-keep”: across sessions:
copies of messages on names of folders and
different clients mappings between
Ì POP3 is stateless message IDs and folder
name
across sessions
2: Application Layer 60
61. DNS: Domain Name System
People: many identifiers: Domain Name System:
SSN, name, passport # Ì distributed database
implemented in hierarchy of
Internet hosts, routers:
many name servers
IP address (32 bit) -
Ì application-layer protocol
used for addressing
host, routers, name servers to
datagrams
communicate to resolve names
“name”, e.g., (address/name translation)
ww.yahoo.com - used by
note: core Internet
humans
function, implemented as
Q: map between IP application-layer protocol
addresses and name ? complexity at network’s
“edge”
2: Application Layer 61
62. DNS
DNS services Why not centralize DNS?
Ì hostname to IP Ì single point of failure
address translation Ì traffic volume
Ì host aliasing Ì distant centralized
Canonical, alias names database
Ì mail server aliasing Ì maintenance
Ì load distribution
replicated Web servers:
doesn’t scale!
set of IP addresses for
one canonical name
2: Application Layer 62
63. Distributed, Hierarchical Database
Root DNS Servers
com DNS servers org DNS servers edu DNS servers
pbs.org poly.edu umass.edu
yahoo.com amazon.com
DNS servers DNS serversDNS servers
DNS servers DNS servers
Client wants IP for www.amazon.com; 1st approx:
Ì client queries a root server to find com DNS server
Ì client queries com DNS server to get amazon.com
DNS server
Ì client queries amazon.com DNS server to get IP
address for www.amazon.com
2: Application Layer 63
64. DNS: Root name servers
Ì contacted by local name server that can not resolve name
Ì root name server:
contacts authoritative name server if name mapping not known
gets mapping
returns mapping to local name server
a Verisign, Dulles, VA
c Cogent, Herndon, VA (also LA)
d U Maryland College Park, MD k RIPE London (also 16 other locations)
g US DoD Vienna, VA
h ARL Aberdeen, MD i Autonomica, Stockholm (plus
j Verisign, ( 21 locations) 28 other locations)
e NASA Mt View, CA m WIDE Tokyo (also Seoul,
f Internet Software C. Palo Alto, Paris, SF)
CA (and 36 other locations)
13 root name
servers worldwide
b USC-ISI Marina del Rey, CA
l ICANN Los Angeles, CA
2: Application Layer 64
65.
66. TLD and Authoritative Servers
Ì Top-level domain (TLD) servers:
responsible for com, org, net, edu, etc, and all
top-level country domains uk, fr, ca, jp.
Network Solutions maintains servers for com TLD
Educause for edu TLD
Ì Authoritative DNS servers:
organization’s DNS servers, providing
authoritative hostname to IP mappings for
organization’s servers (e.g., Web, mail).
can be maintained by organization or service
provider
2: Application Layer 66
67. Local Name Server
Ì does not strictly belong to hierarchy
Ì each ISP (residential ISP, company,
university) has one.
also called “default name server”
Ì when host makes DNS query, query is sent
to its local DNS server
acts as proxy, forwards query into hierarchy
2: Application Layer 67
68. DNS name root DNS server
resolution example
2
Ì Host at cis.poly.edu 3
TLD DNS server
wants IP address for 4
gaia.cs.umass.edu 5
iterated query: local DNS server
dns.poly.edu
Ì contacted server 6
7
replies with name of 1 8
server to contact
Ì “I don’t know this authoritative DNS server
dns.cs.umass.edu
name, but ask this requesting host
server” cis.poly.edu
gaia.cs.umass.edu
2: Application Layer 68
69. DNS name
resolution example root DNS server
recursive query: 2 3
Ì puts burden of name
7 6
resolution on
TLD DNS server
contacted name
server
Ì heavy load? local DNS server
dns.poly.edu 5 4
1 8
authoritative DNS server
dns.cs.umass.edu
requesting host
cis.poly.edu
gaia.cs.umass.edu
2: Application Layer 69
70. DNS: caching and updating records
Ì once (any) name server learns mapping, it caches
mapping
cache entries timeout (disappear) after some
time
TLD servers typically cached in local name
servers
• Thus root name servers not often visited
Ì update/notify mechanisms under design by IETF
RFC 2136
http://www.ietf.org/html.charters/dnsind-charter.html
2: Application Layer 70
71. DNS records
DNS: distributed db storing resource records (RR)
RR format: (name, value, type, ttl)
Ì Type=A Ì Type=CNAME
name is hostname name is alias name for some
value is IP address “canonical” (the real) name
Ì Type=NS www.ibm.com is really
servereast.backup2.ibm.com
name is domain (e.g.
value is canonical name
foo.com)
value is hostname of Ì Type=MX
authoritative name server
value is name of mailserver
for this domain
associated with name
2: Application Layer 71
72. DNS protocol, messages
DNS protocol : query and reply messages, both with
same message format
msg header
Ì identification: 16 bit #
for query, reply to query
uses same #
Ì flags:
query or reply
recursion desired
recursion available
reply is authoritative
2: Application Layer 72
73. DNS protocol, messages
Name, type fields
for a query
RRs in response
to query
records for
authoritative servers
additional “helpful”
info that may be used
2: Application Layer 73
74. Inserting records into DNS
Ì example: new startup “Network Utopia”
Ì register name networkuptopia.com at DNS registrar
(e.g., Network Solutions)
provide names, IP addresses of authoritative name server
(primary and secondary)
registrar inserts two RRs into com TLD server:
(networkutopia.com, dns1.networkutopia.com, NS)
(dns1.networkutopia.com, 212.212.212.1, A)
Ì create authoritative server Type A record for
www.networkuptopia.com; Type MX record for
networkutopia.com
Ì How do people get IP address of your Web site?
2: Application Layer 74
75. Pure P2P architecture
Ì no always-on server
Ì arbitrary end systems
directly communicate peer-peer
Ì peers are intermittently
connected and change IP
addresses
Ì Three topics:
File distribution
Searching for information
– Case Study: Skype/Bit-
torrent
2: Application Layer 75
76. File Distribution: Server-Client vs P2P
Question : How much time to distribute file
from one server to N peers?
us: server upload
Server
bandwidth
ui: peer i upload
u1 d1 u2 bandwidth
us d2
di: peer i download
File, size F bandwidth
dN
Network (with
uN abundant bandwidth)
2: Application Layer 76
77. File distribution time: server-client
Server
Ì server sequentially F u1 d1 u2
sends N copies: us d2
NF/us time dN Network (with
abundant bandwidth)
Ì client i takes F/di uN
time to download
Time to distribute F
to N clients using = dcs = max { NF/us, F/min(di) }
i
client/server approach
increases linearly in N
(for large N) 2: Application Layer 77
78. File distribution time: P2P
Server
Ì server must send one
F u1 d1 u2
copy: F/us time us d2
Ì client i takes F/di time Network (with
dN
to download abundant bandwidth)
uN
Ì NF bits must be
downloaded (aggregate)
Ì fastest possible upload rate: u +
s Σ ui
dP2P = max { F/us, F/min(di) , NF/(us + Σui) }
i
2: Application Layer 78
79. Server-client vs. P2P: example
Client upload rate = u, F/u = 1 hour, us = 10u, dmin ≥ us
3.5
P2P
Minimum Distribution Time
3
Client-Server
2.5
2
1.5
1
0.5
0
0 5 10 15 20 25 30 35
N
2: Application Layer 79
80. Types of P2P Architectures
Centralized
--E.g., Napster
Pure
--E.g., Gnutella
Hybrid
--E.g., Bit-torrent
81. Key Differences
Centralized
-A server keeps track of location of files &
corresponding peers
-Peer only needs to contact the server to get
the download location
-Insecure due to single point of failure
Pure
-No tracking of files
-Queries are flooded throughout the network
-Robust as there is no single point of failure
82. Key Differences
Hybrid
-Overlay networks –logically connected nodes
-Publish file locations on websites/blogs etc
-A group of “trackers” maintain file locations
(Bit-torrent for example)
-Distributed Hash Tables are used for
efficient searching of file locations
Useful Links
http://bittorrent.org/bittorrentecon.pdf
http://dessent.net/btfaq/
83. File distribution: BitTorrent
Ì P2P file distribution
tracker: tracks peers torrent: group of
participating in torrent peers exchanging
chunks of a file
obtain list
of peers
trading
chunks
peer
2: Application Layer 83
84. BitTorrent (1)
Ì file divided into 256KB chunks.
Ì peer joining torrent:
has no chunks, but will accumulate them over time
registers with tracker to get list of peers,
connects to subset of peers (“neighbors”)
Ì while downloading, peer uploads chunks to other
peers.
Ì peers may come and go
Ì once peer has entire file, it may (selfishly) leave or
(altruistically) remain
2: Application Layer 84
85. BitTorrent (2) Sending Chunks: tit-for-tat
Ì Alice sends chunks to four
Pulling Chunks
neighbors currently
Ì at any given time,
sending her chunks at the
different peers have highest rate
different subsets of
re-evaluate top 4 every
file chunks
10 secs
Ì periodically, a peer
Ì every 30 secs: randomly
(Alice) asks each
neighbor for list of select another peer,
chunks that they have. starts sending chunks
newly chosen peer may
Ì Alice sends requests
for her missing chunks join top 4
rarest first
“optimistically unchoke”
2: Application Layer 85
86. BitTorrent: Tit-for-tat
(1) Alice “optimistically unchokes” Bob
(2) Alice becomes one of Bob’s top-four providers; Bob reciprocates
(3) Bob becomes one of Alice’s top-four providers
With higher upload rate,
can find better trading
partners & get file faster!
2: Application Layer 86
87. Distributed Hash Table (DHT)
Ì DHT = distributed P2P database
Ì Database has (key, value) pairs;
key: ss number; value: human name
key: content type; value: IP address
Ì Peers query DB with key
DB returns values that match the key
Ì Peers can also insert (key, value) peers
88. DHT Identifiers
Ì Assign integer identifier to each peer in range
[0,2n-1].
Each identifier can be represented by n bits.
Ì Require each key to be an integer in same range.
Ì To get integer keys, hash original key.
eg, key = h(“Led Zeppelin IV”)
This is why they call it a distributed “hash” table
89. How to assign keys to peers?
Ì Central issue:
Assigning (key, value) pairs to peers.
Ì Rule: assign key to the peer that has the
closest ID.
Ì Convention in lecture: closest is the
immediate successor of the key.
Ì Ex: n=4; peers: 1,3,4,5,8,10,12,14;
key = 13, then successor peer = 14
key = 15, then successor peer = 1
90. Circular DHT (1)
1
15 3
4
12
5
10
8
Ì Each peer only aware of immediate successor
and predecessor.
Ì “Overlay network”
91. Circle DHT (2)
O(N) messages 0001 Who’s resp
on avg to resolve for key 1110 ?
I am
query, when there 0011
are N peers
1111
1110
1110
0100
1110
1100
1110
1110 0101
Define closest 1110
as closest 1010
successor 1000
92. Circular DHT with Shortcuts
1 Who’s resp
for key 1110?
3
15
4
12
5
10
8
Ì Each peer keeps track of IP addresses of predecessor,
successor, short cuts.
Ì Reduced from 6 to 2 messages.
Ì Possible to design shortcuts so O(log N) neighbors, O(log
N) messages in query
93. Peer Churn
1
•To handle peer churn, require
3 each peer to know the IP address
15
of its two successors.
• Each peer periodically pings its
4
two successors to see if they
12 are still alive.
5
10
8
Ì Peer 5 abruptly leaves
Ì Peer 4 detects; makes 8 its immediate successor;
asks 8 who its immediate successor is; makes 8’s
immediate successor its second successor.
Ì What if peer 13 wants to join?
94. P2P Case study: Skype
Skype clients (SC)
Ì inherently P2P: pairs
of users communicate.
Ì proprietary Skype
application-layer login server Supernode
protocol (inferred via (SN)
reverse engineering)
Ì hierarchical overlay
with SNs
Ì Index maps usernames
to IP addresses;
distributed over SNs
2: Application Layer 94
95. Peers as relays
Ì Problem when both
Alice and Bob are
behind “NATs”.
NAT prevents an outside
peer from initiating a call
to insider peer
Ì Solution:
Using Alice’s and Bob’s
SNs, Relay is chosen
Each peer initiates
session with relay.
Peers can now
communicate through
NATs via relay
2: Application Layer 95
96. Key Differences Re-visited
-Client-Server architecture poor scaling for
Server
-Pure/Centralized P2P architecture good
scaling but performs poorly when upload
speeds of peer is bad (due to individual TCP
connections
-Hybrid/Torrents –advantage of both
centralized/pure P2P by splitting files into
multiple chunks and off-loading work to
different peers (creating more sources)
97. Chapter 2: Summary
our study of network apps now complete!
Ì application architectures Ì specific protocols:
client-server HTTP
P2P FTP
hybrid SMTP, POP, IMAP
DNS
Ì application service
P2P: BitTorrent, Skype
requirements:
reliability, bandwidth, Ì socket programming
delay
Ì Internet transport
service model
connection-oriented,
reliable: TCP
unreliable, datagrams: UDP
2: Application Layer 97
98. Chapter 2: Summary
Most importantly: learned about protocols
Ì typical request/reply Important themes:
message exchange: Ì control vs. data msgs
client requests info or
in-band, out-of-band
service
server responds with Ì centralized vs.
data, status code decentralized
Ì message formats: Ì stateless vs. stateful
headers: fields giving Ì reliable vs. unreliable
info about data
msg transfer
data: info being
Ì “complexity at network
communicated
edge”
2: Application Layer 98