Application layer: Roadmap Principles of network applications Web - PowerPoint PPT Presentation

Application Layer Application layer: Roadmap  Principles of network applications  Web and HTTP  FTP  Electronic Mail  SMTP, POP3, IMAP  DNS  P2P applications  Socket programming with UDP  Socket programming with TCP 79 CSE Department

Application Layer Pure P2P architecture  no always-on server  arbitrary end systems directly communicate peer-peer  peers are intermittently connected and change IP addresses  Topics:  File distribution  Searching for information 80 CSE Department

Application Layer Peer-2-Peer Architectures  Unstructured  Centralized: Napster  Decentralized: Gnutella  Hierarchical : KaZaA  Structured  DHT-based 81 CSE Department

Application Layer File Distribution: Server-Client vs P2P Question : How much time to distribute file from one server to N peers ? u s : server upload bandwidth Server u i : peer i upload bandwidth u 2 u 1 d 1 d 2 u s d i : peer i download File, size F bandwidth d N Network (with abundant bandwidth) u N 82 CSE Department

Application Layer File distribution time: server-client Server  server sequentially u 2 F u 1 d 1 sends N copies: d 2 u s  NF/u s time Network (with d N  client i takes F/d i abundant bandwidth) u N time to download Time to distribute F to N clients using = d cs = max { NF/u s , F/min(d i ) } client/server approach i increases linearly in N (for large N) 83 CSE Department

Application Layer File distribution time: P2P Server  server must send one u 2 F copy: F /u s time u 1 d 1 d 2 u s  client i takes F/d i time Network (with to download d N abundant bandwidth)  NF bits must be u N downloaded (aggregate)  fastest possible upload rate: u s + S u i d P2P = max { F/u s , F/min(d i ) , NF/(u s + S u i ) } i 84 CSE Department

Application Layer Server-client vs. P2P: example Client upload rate = u, F/u = 1 hour, u s = 10u, d min ≥ u s 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 85 CSE Department

Application Layer File distribution: BitTorrent  P2P file distribution torrent: group of tracker: tracks peers peers exchanging participating in torrent chunks of a file obtain list of peers trading chunks Bram Cohen peer 86 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 87 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 88 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 89 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 90 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 91 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 92 CSE Department

Application Layer Overall Architecture Tracker Web Server C A Peer Peer [Seed] B [Leech] Downloader Peer “US” [Leech] 93 CSE Department

Application Layer BitTorrent (1)  file divided into 256KB chunks .  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”)  while downloading, peer uploads chunks to other peers.  peers may come and go  once peer has entire file, it may (selfishly) leave or (altruistically) remain 94 CSE Department

Application Layer BitTorrent (2) Sending Chunks: tit-for-tat  Alice sends chunks to four Pulling Chunks neighbors currently  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  periodically, a peer (Alice) asks each  every 30 secs: randomly neighbor for list of select another peer, chunks that they have. starts sending chunks  newly chosen peer may  Alice sends requests join top 4 for her missing chunks  rarest first 95 CSE Department

Application Layer Hash Table 96 CSE Department

Application Layer Distributed Hash Table (DHT)  DHT = distributed P2P database  Database has (key, value) pairs;  key: ss number; value: human name  key: content type; value: IP address  Peers query DB with key  DB returns values that match the key  Peers can also insert (key, value) 97 CSE Department

Application Layer DHT Identifiers  Assign integer identifier to each peer in range [0,2 n -1].  Each identifier can be represented by n bits.  Require each key to be an integer in same range.  To get integer keys , hash original key.  eg, key = h(“Led Zeppelin IV”)  This is why they call it a distributed “hash” table 98 CSE Department

Application Layer How to assign keys to peers?  Central issue:  Assigning (key, value) pairs to peers.  Rule: assign key to the peer that has the closest ID.  Convention in lecture: closest is the immediate successor of the key.  Ex: n=4; peers: 1,3,4,5,8,10,12,14;  key = 13, then successor peer = 14  key = 15, then successor peer = 1 99 CSE Department

Application Layer Circular DHT (1) Overlay Link 1 Not the real link! 3 15 4 12 5 10 8  Each peer only aware of immediate successor and predecessor.  “ Overlay network ” 100 CSE Department

Application Layer Circle DHT (2) O(N) messages Who’s resp 0001 on avg to resolve for key 1110 ? query, when there I am are N peers 0011 1111 1110 0100 1110 1110 1100 0101 1110 1110 Define closest 1110 as closest 1010 1000 successor 101 CSE Department

Application Layer Circular DHT with Shortcuts 1 Who’s resp for key 1110? 3 15 4 12 5 How to achieve O(log N) search time? 10 8  Each peer keeps track of IP addresses of predecessor, successor, short cuts .  Reduced from 6 to 2 messages.  Possible to design shortcuts so O(log N) neighbors, O(log N) messages in query 102 CSE Department

Application Layer Peer Churn 1 • To handle peer churn, require each peer to know the IP address 3 of its two successors. 15 • Each peer periodically pings its two successors to see if they 4 are still alive . 12 5 10 8  Peer 5 abruptly leaves  Peer 4 detects; makes 8 its immediate successor; asks 8 who its immediate successor is; makes 8’s immediate successor its second successor.  What if peer 13 wants to join? 103 CSE Department

Application Layer Chapter 2: Summary our study of network apps now complete!  specific protocols:  application architectures  HTTP  client-server  FTP  P2P  SMTP, POP, IMAP  hybrid  DNS  P2P: BitTorrent, Skype  application service  socket programming requirements:  reliability, bandwidth, delay  Internet transport service model  connection-oriented, reliable: TCP  unreliable, datagrams: UDP 127 CSE Department

Application Layer Chapter 2: Summary Most importantly: learned about protocols  typical request/reply Important themes: message exchange:  control vs. data msgs  in-band, out-of-band  client requests info or service  centralized vs.  server responds with decentralized data, status code  stateless vs. stateful  message formats:  reliable vs. unreliable  headers: fields giving msg transfer info about data  “complexity at network  data: info being edge” communicated 128 CSE Department