peer-to-peer protocols and systems ta: david murray 15-441 spring 2006 4/19/2006
TRANSCRIPT
![Page 1: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/1.jpg)
Peer-to-PeerProtocols and Systems
TA: David Murray
15-441 Spring 2006
4/19/2006
![Page 2: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/2.jpg)
2
P2P - Outline
• What is P2P?
• P2P System Types– 1) File-sharing– 2) File distribution– 3) Streaming
• Uses & Challenges
![Page 3: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/3.jpg)
3
Problem: Scalability
• Hundreds of clients => 1 server– OK
• Thousands of clients => 1 server– Maybe OK
• Millions/billions of clients => 1 server– What happens?...
![Page 4: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/4.jpg)
4
![Page 5: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/5.jpg)
5
Solution:Distribute the cost among the end users
![Page 6: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/6.jpg)
6
Three Classes of P2P Systems
• 1) File-sharing– (old) Napster (centralized)– Gnutella (flooding)– KaZaA (intelligent flooding)– DHTs/Chord (structured overlay routing)
• 2) File distribution– BitTorrent
• 3) Streaming– End System Multicast (a.k.a. Overlay Multicast)
![Page 7: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/7.jpg)
1) P2P File-sharing Systems
![Page 8: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/8.jpg)
8
1) P2P File-sharing Systems
• Centralized Database– (old) Napster
• Query Flooding– Gnutella
• Intelligent Query Flooding– KaZaA
• Structured Overlay Routing– Distributed Hash Tables
![Page 9: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/9.jpg)
9
File searching
Internet
N1
N2 N3
N6N5
N4
Publisher
Key=“title”Value=MP3 data… Client
Lookup(“title”)
?
![Page 10: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/10.jpg)
10
File searching
• Needles vs. Haystacks– Searching for top 40, or an obscure punk
track from 1981 that nobody’s heard of?
• Search expressiveness– Whole word? Regular expressions? File
names? Attributes? Whole-text search?• (e.g., p2p gnutella or p2p google?)
![Page 11: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/11.jpg)
11
File-sharing: Framework
• Common Primitives:– Join: how do I begin participating?– Publish: how do I advertise my file?– Search: how do I find a file?– Fetch: how do I retrieve a file?
![Page 12: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/12.jpg)
12
P2P File-sharing Systems
• Centralized Database– (old) Napster
• Query Flooding– Gnutella
• Intelligent Query Flooding– KaZaA
• Structured Overlay Routing– Distributed Hash Tables
![Page 13: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/13.jpg)
13
(old) Napster: History
• 1999: Sean Fanning launches Napster
• Peaked at 1.5 million simultaneous users
• Jul 2001: Napster shuts down
• [2003: Napster’s name reused for an online music service (no relation)]
![Page 14: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/14.jpg)
14
(old) Napster: Overview
• Centralized Database:– Join: on startup, client contacts central
server– Publish: reports list of files to central
server– Search: query the server => return
someone that stores the requested file– Fetch: get the file directly from peer
![Page 15: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/15.jpg)
15
Napster: Publish
I have lots of TV show theme song files!
Publish
insert(XenaThemeSong.mp3, 123.2.21.23)Insert(HerculesThemeSong.mp3, 123.2.21.23)...
123.2.21.23
![Page 16: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/16.jpg)
16
Napster: Search
“Where is the XenaTheme song?”
Reply
search(Xena)Fetch
Query
![Page 17: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/17.jpg)
17
Napster: Discussion
• Pros:– Simple– Search scope is O(1)– Controllable (pro or con?)
• Cons:– Server maintains O(N) State– Server does all processing– Single point of failure
![Page 18: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/18.jpg)
18
P2P File-sharing Systems
• Centralized Database– (old) Napster
• Query Flooding– Gnutella
• Intelligent Query Flooding– KaZaA
• Structured Overlay Routing– Distributed Hash Tables
![Page 19: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/19.jpg)
19
Gnutella: History
• In 2000, J. Frankel and T. Pepper from Nullsoft released Gnutella
• Soon many other clients: Bearshare, Morpheus, LimeWire, etc.
• In 2001, many protocol enhancements including “ultrapeers”
![Page 20: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/20.jpg)
20
Gnutella: Overview
• Query Flooding:– Join: on startup, client contacts a few other
nodes; these become its “neighbors”– Publish: (no need)– Search: ask neighbors, who ask their
neighbors, and so on... when/if found, reply to sender.
• TTL limits propagation– Fetch: get the file directly from peer
![Page 21: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/21.jpg)
21
“I have XenaEpisode1.mpg”“I haveXenaEpisode1.mpg”
Gnutella: Search
“Where isXenaEpisode1.mpg?”
Query
Reply
![Page 22: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/22.jpg)
22
Gnutella: Discussion• Pros:
– Fully de-centralized– Search cost distributed– Processing @ each node permits powerful search
semantics
• Cons:– Search scope is O(N)– Search time is O(???)– Nodes leave often, network unstable
• TTL-limited search works well for haystacks.– For scalability, does NOT search every node. May have to
re-issue query later
![Page 23: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/23.jpg)
23
P2P File-sharing Systems
• Centralized Database– (old) Napster
• Query Flooding– Gnutella
• Intelligent Query Flooding– KaZaA
• Structured Overlay Routing– Distributed Hash Tables
![Page 24: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/24.jpg)
24
KaZaA: History
• In 2001, KaZaA created by Dutch company Kazaa BV
• Single network called FastTrack used by other clients as well: Morpheus, giFT, etc.
• Eventually protocol changed so other clients could no longer talk to it
![Page 25: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/25.jpg)
25
KaZaA: Overview
• “Smart” Query Flooding:– Join: on startup, client contacts a “supernode” ...
may at some point become one itself– Publish: send list of files to supernode– Search: send query to supernode, supernodes
flood query amongst themselves.– Fetch: get the file directly from peer(s); can fetch
simultaneously from multiple peers
![Page 26: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/26.jpg)
26
KaZaA: Network Design“Super Nodes”
![Page 27: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/27.jpg)
27
KaZaA: File Insert
I have lots of files!
Publishinsert(Xena.mp3, 123.2.21.23)Insert(Hercules.mp3,123.2.21.23)
123.2.21.23
![Page 28: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/28.jpg)
28
KaZaA: File Search
Where is Xena.mp3?
Query
search(Xena.mp3)-->123.2.0.18Replies
search(Xena.mp3)-->123.2.22.50
123.2.0.18
123.2.22.50
![Page 29: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/29.jpg)
29
KaZaA: Fetching
• More than one node may have requested file...• How to tell?
– Must be able to distinguish identical files– Not necessarily same filename– Same filename not necessarily same file...
• Use Hash of file– KaZaA uses its own “UUHash”: fast, but not secure– Alternatives: MD5, SHA-1
• How to fetch?– Get bytes [0..1000] from A, [1001...2000] from B– Alternative: Erasure Codes
![Page 30: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/30.jpg)
30
KaZaA: Discussion
• Pros:– Tries to take into account node heterogeneity:
• Bandwidth
• Host Computational Resources
• Host Availability (?)
– Rumored to take into account network locality
• Cons:– Mechanisms easy to circumvent– Still no real guarantees on search scope or search time
• Similar behavior to Gnutella, but better.
![Page 31: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/31.jpg)
31
Stability and Superpeers
• Why supernodes?– Query consolidation
• Many connected nodes may have only a few files• Propagating a query to a sub-node would take more b/w
than answering it yourself
– Caching effect• Requires network stability
• Supernode selection is time-based– How long you’ve been on is a good predictor of
how long you’ll be around.
![Page 32: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/32.jpg)
32
P2P File-sharing Systems
• Centralized Database– (old) Napster
• Query Flooding– Gnutella
• Intelligent Query Flooding– KaZaA
• Structured Overlay Routing– Distributed Hash Tables
![Page 33: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/33.jpg)
33
Distributed Hash Tables
• Academic answer to P2P• Goals
– Guaranteed lookup success– Provable bounds on search time– Provable scalability
• Makes some things harder– Fuzzy queries / full-text search / etc.
• Read-write, not read-only• Hot Topic in networking since introduction in
~2000/2001
![Page 34: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/34.jpg)
34
DHT: Overview
• Abstraction: a distributed “hash-table” (DHT) data structure:
– put(id, item);– item = get(id);
• Implementation: nodes in system form a distributed data structure
– Can be Ring, Tree, Hypercube, Skip List, Butterfly Network, ...
![Page 35: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/35.jpg)
35
DHT: Overview (continued)
• Structured Overlay Routing:– Join: On startup, contact a “bootstrap” node and integrate
yourself into the distributed data structure; get a node id– Publish: Route publication for file id toward a close node id
along the data structure– Search: Route a query for file id toward a close node id.
Data structure guarantees that query will meet the publication. (Note: cannot do keyword search)
– Fetch: Two options:• Publication contains actual file => fetch from where query stops• Publication says “I have file X” => query tells you 128.2.1.3 has
X, use IP routing to get X from 128.2.1.3
![Page 36: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/36.jpg)
36
DHT: Example – Chord
• Associate to each node and file a unique id in an uni-dimensional space (a Ring)
– E.g., pick from the range [0...2m]– Usually the hash of the file or IP address
• Properties:– “It allows a distributed set of participants to agree
on a single node as a rendezvous point for a given key, without any central coordination.” (Chord site)
– Can find data using O(log N) messages, where N is the total number of nodes
• (Why? We’ll see…)
from MIT in 2001
![Page 37: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/37.jpg)
37
DHT: Consistent Hashing
N32
N90
N105
K80
K20
K5
Circular ID space
Key 5Node 105
A key is stored at its successor: node with next higher ID
![Page 38: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/38.jpg)
38
DHT: Chord Basic Lookup
N32
N90
N105
N60
N10N120
K80
“Where is key 80?”
“N90 has K80”
![Page 39: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/39.jpg)
39
DHT: Chord “Finger Table”
N80
1/21/4
1/8
1/161/321/641/128
• Entry i in the finger table of node n is the first node that succeeds or equals n + 2i
• In other words, the ith finger points 1/2n-i way around the ring• (This is what makes O(log N) messages for a retrieval possible!)
![Page 40: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/40.jpg)
40
DHT: Chord Join
• Assume an identifier space [0..8]
• Node n1 joins0
1
2
34
5
6
7
i id+2i succ0 2 11 3 12 5 1
Succ. Table
![Page 41: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/41.jpg)
41
DHT: Chord Join
• Node n2 joins0
1
2
34
5
6
7
i id+2i succ0 2 21 3 12 5 1
Succ. Table
i id+2i succ0 3 11 4 12 6 1
Succ. Table
![Page 42: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/42.jpg)
42
DHT: Chord Join
• Nodes n0, n6 join
01
2
34
5
6
7
i id+2i succ0 2 21 3 62 5 6
Succ. Table
i id+2i succ0 3 61 4 62 6 6
Succ. Table
i id+2i succ0 1 11 2 22 4 0
Succ. Table
i id+2i succ0 7 01 0 02 2 2
Succ. Table
![Page 43: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/43.jpg)
43
DHT: Chord Join
• Nodes: n1, n2, n0, n6
• Items: f7, f2
01
2
34
5
6
7 i id+2i succ0 2 21 3 62 5 6
Succ. Table
i id+2i succ0 3 61 4 62 6 6
Succ. Table
i id+2i succ0 1 11 2 22 4 0
Succ. Table
7
Items 1
Items
i id+2i succ0 7 01 0 02 2 2
Succ. Table
![Page 44: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/44.jpg)
44
DHT: Chord Routing
• Upon receiving a query for item id, a node:
• Checks whether stores the item locally
• If not, forwards the query to the largest node in its successor table that does not exceed id
01
2
34
5
6
7 i id+2i succ0 2 21 3 62 5 6
Succ. Table
i id+2i succ0 3 61 4 62 6 6
Succ. Table
i id+2i succ0 1 11 2 22 4 0
Succ. Table
7
Items 1
Items
i id+2i succ0 7 01 0 02 2 2
Succ. Table
query(7)
![Page 45: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/45.jpg)
45
DHT: Chord Summary
• Routing table size?–Log N fingers
• Routing time?–Each hop expects to 1/2 the distance to the
desired id => expect O(log N) hops.
![Page 46: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/46.jpg)
46
DHT: Discussion
• Pros:– Guaranteed Lookup– O(log N) per node state and search scope
• Cons:– Does *not* support keyword search – No one uses them? (only one file sharing
app)– Supporting non-exact match search is hard
![Page 47: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/47.jpg)
47
1) P2P File-sharing: Summary
• Many different styles; remember pros and cons of each– centralized, flooding, intelligent flooding, overlay routing
• Lessons learned:– Single points of failure are very bad– Flooding messages to everyone is bad– Underlying network topology is important– Not all nodes need be equal– Need incentives to discourage freeloading– Privacy and security matter– Structure can provide theoretical bounds and guarantees
![Page 48: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/48.jpg)
2) P2P File Distribution Systems
(i.e. BitTorrent)
![Page 49: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/49.jpg)
49
BitTorrent: History
• In 2002, B. Cohen debuted BitTorrent• Key Motivation:
– Popularity exhibits temporal locality (Flash Crowds)– E.g., Slashdot effect, CNN on 9/11, new movie/game
release
• Focused on Efficient Fetching, not Searching:– Distribute the same file to all peers– Single publisher, multiple downloaders
• Has many “real” publishers:– Example: Blizzard Entertainment uses it for World of
Warcraft update distribution
![Page 50: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/50.jpg)
50
BitTorrent: Overview• Focused in efficient fetching, not searching• Swarming:
– Join: contact centralized “tracker” server, get a list of peers.
– Publish: Run a tracker server.– Search: n/a (need to find elsewhere, i.e. Google)– Fetch: Download chunks of the file from your peers.
Upload chunks you have to them.
• Improvements from old Napster-era:– Chunk based downloading; few large files– Anti-freeloading mechanisms
![Page 51: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/51.jpg)
51
BitTorrent: Publish/JoinTracker
![Page 52: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/52.jpg)
52
BitTorrent: Fetch
![Page 53: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/53.jpg)
53
BitTorrent: Summary
• Pros:– Works reasonably well in practice– Gives peers incentive to share resources;
avoids freeloaders
• Cons:– No search; only content distribution– Central tracker server needed to bootstrap
swarm
![Page 54: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/54.jpg)
3) P2P Streaming Systems
(i.e. Overlay Multicast a.k.a.End System Multicast)
![Page 55: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/55.jpg)
55
Live Broadcast: Pre-Internet
• Tower/Cable/Satellite TV, Radio
• Problems– Limitations
• # of channels• Physical reach
– Cost• Content production monopolized by big corps.• Content distribution monopolized by big corps.
![Page 56: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/56.jpg)
56
Live Internet Broadcast: Pre-P2P
• Unicast streaming– Small-scale video conferencing– Large-scale streaming (AOL Live, etc.)
• Problems– Unicast streaming requires lots of
bandwidth• Example: AOL webcast of Live 8 concert, July
2, 2005: 1500 servers in 90 locations = $$$
![Page 57: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/57.jpg)
57
Solution...?Use IP Multicast!...?
![Page 58: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/58.jpg)
58
Solution: IP Multicast?
• On a single LAN: GREAT!– Can distribute traffic to everyone at once without
duplicating streams, set TTL=1, no problem!
• Cross-LAN…Problem – Requires multicast-enabled routers
• Must allocate resources toward processing multicast packets
• As a result, *MANY, MANY* computers can’t receive IP multicast packets from outside their LAN
![Page 59: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/59.jpg)
59
Solution (again!):Don’t depend on routers…
Distribute the cost among the end users
![Page 60: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/60.jpg)
60
Internet broadcasting Structure
RouterSourceApplication end-point
![Page 61: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/61.jpg)
61
End System Multicast (ESM) (a.k.a Overlay Multicast)
+ Instantly deployable+ Enables ubiquitous broadcast
RouterSourceApplication end-point
![Page 62: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/62.jpg)
62
Example ESM Treehttp://esm.cs.cmu.edu
![Page 63: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/63.jpg)
63
Single Overlay Distribution Tree
B
C
E
F
D
A
G
![Page 64: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/64.jpg)
64
Single Overlay Distribution Tree
B
C
E
F
D
A
G
IN: n kb/sec
OUT: 2n kb/sec
![Page 65: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/65.jpg)
65
Multiple Overlay Distribution Trees
B
C
E
F
D
A
G
![Page 66: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/66.jpg)
66
Multiple Overlay Distribution Trees
B
C
E
F
D
A
G
![Page 67: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/67.jpg)
67
Multiple Overlay Distribution Trees
B
C
E
F
D
A
G
![Page 68: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/68.jpg)
68
Multiple Overlay Distribution Trees
B
C
EF
DA
G D
E G
B A F C
A
B C
F G D E
N kb/sec
N/2 kb/sec N/2 kb/sec
![Page 69: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/69.jpg)
69
My Research with ESM
• Can we combine the best parts of multicast with ESM?– My solution:
• Integrate LAN Multicast (i.e. IP Multicast with TTL=1) with ESM
• Each LAN has 1 or more forwarders from the outside receiving data which gets forwarded on multicast with TTL=1
• “Nodes” of overlay trees are now LANs instead of individual hosts
![Page 70: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/70.jpg)
70
ESM + LAN Multicast
LAN B
LAN C
LAN E
LAN F
LAN D
LAN A
LAN G
![Page 71: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/71.jpg)
71
P2P Systems: Summary
• 3 types of P2P systems– File-sharing
• Centralized ((old) Napster), Flooding (Gnutella), Intelligent Flooding (KaZaA) Overlay Routing (DHTs/Chord)
– File distribution• BitTorrent
– Streaming• End System Multicast a.k.a. Overlay Multicast
• Lessons– Single points of failure are very bad– Underlying network topology is important– Not all nodes are equal– Can’t depend on routers to satisfy all of your networking desires
• Room for growth– Privacy & Security– Research is ongoing
![Page 72: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/72.jpg)
Extra Slides(From previous P2P lectures)
![Page 73: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/73.jpg)
73
KaZaA: Usage Patterns• KaZaA is more than
one workload!– Many files < 10MB
(e.g., Audio Files)– Many files > 100MB
(e.g., Movies)
from Gummadi et al., SOSP 2003
![Page 74: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/74.jpg)
74
KaZaA: Usage Patterns (2)• KaZaA is not Zipf!
– FileSharing: “Request-once”
– Web: “Request-repeatedly”
from Gummadi et al., SOSP 2003
![Page 75: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/75.jpg)
75
KaZaA: Usage Patterns (3)
• What we saw:– A few big files consume most of the bandwidth– Many files are fetched once per client but still very popular
• Solution?– Caching!
from Gummadi et al., SOSP 2003
![Page 76: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/76.jpg)
76
Freenet: History
• In 1999, I. Clarke started the Freenet project
• Basic Idea:– Employ Internet-like routing on the overlay
network to publish and locate files
• Addition goals:– Provide anonymity and security– Make censorship difficult
![Page 77: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/77.jpg)
77
Freenet: Overview
• Routed Queries:– Join: on startup, client contacts a few other
nodes it knows about; gets a unique node id– Publish: route file contents toward the file id. File
is stored at node with id closest to file id– Search: route query for file id toward the closest
node id– Fetch: when query reaches a node containing file
id, it returns the file to the sender
![Page 78: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/78.jpg)
78
Freenet: Routing Tables• id – file identifier (e.g., hash of file)• next_hop – another node that stores the file id• file – file identified by id being stored on the local node
• Forwarding of query for file id– If file id stored locally, then stop
• Forward data back to upstream requestor– If not, search for the “closest” id in the table, and
forward the message to the corresponding next_hop
– If data is not found, failure is reported back• Requestor then tries next closest match in routing
table
id next_hop file
……
![Page 79: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/79.jpg)
79
Freenet: Routing
4 n1 f412 n2 f12 5 n3
9 n3 f9
3 n1 f314 n4 f14 5 n3
14 n5 f1413 n2 f13 3 n6
n1 n2
n3
n4
4 n1 f410 n5 f10 8 n6
n5
query(10)
1
2
3
4
4’
5
![Page 80: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/80.jpg)
80
Freenet: Routing Properties
• “Close” file ids tend to be stored on the same node– Why? Publications of similar file ids route toward
the same place
• Network tend to be a “small world”– Small number of nodes have large number of
neighbors (i.e., ~ “six-degrees of separation”)
• Consequence:– Most queries only traverse a small number of hops
to find the file
![Page 81: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/81.jpg)
81
Freenet: Anonymity & Security
• Anonymity– Randomly modify source of packet as it traverses the
network– Can use “mix-nets” or onion-routing
• Security & Censorship resistance– No constraints on how to choose ids for files => easy to
have to files collide, creating “denial of service” (censorship)– Solution: have a id type that requires a private key signature
that is verified when updating the file– Cache file on the reverse path of queries/publications =>
attempt to “replace” file with bogus data will just cause the file to be replicated more!
![Page 82: Peer-to-Peer Protocols and Systems TA: David Murray 15-441 Spring 2006 4/19/2006](https://reader035.vdocuments.site/reader035/viewer/2022062803/56649f3a5503460f94c578e9/html5/thumbnails/82.jpg)
82
Freenet: Discussion
• Pros:– Intelligent routing makes queries relatively short – Search scope small (only nodes along search path
involved); no flooding– Anonymity properties may give you “plausible
deniability”
• Cons:– Still no provable guarantees!– Anonymity features make it hard to measure,
debug