Network Application. Topic. Principle of Network Application. Principle of Network Application

Transcription

1 Topic Network Application o Principle of Network Application o Web & HTTP o FTP o o DNS o Peer-to-peer 2 Principle of Network Application o Network Applications web instant messaging remote login P2P file sharing multi- network games streaming stored video clips voice over IP real-time video conferencing grid computing Etc. Principle of Network Application o Application Architecture Client- Peer-to-peer (P2P) Hybrid of client- and P2P 3 4

2 Client-Server o Architecture client/ : always-on host permanent IP address farms for scaling clients: communicate with may be intermittently connected may have dynamic IP addresses do not communicate directly with each other Peer-to-peer o Architecture peer-peer Peers no always-on arbitrary end systems directly communicate peers are intermittently connected and change IP addresses example: Gnutella Highly scalable but difficult to manage Hybrid of client- and P2P o Skype voice-over-ip P2P application centralized : finding address of remote party: client-client connection: direct (not through ) o Instant messaging chatting between two s is P2P centralized service: client presence detection/location registers its IP address with central when it comes online contacts central to find IP addresses of buddies 8 Processes communicating Process: program running within a host. o within same host, two processes communicate using inter-process communication (defined by OS). o processes in different hosts communicate by exchanging messages Client process: process that initiates communication Server process: process that waits to be contacted o Note: applications with P2P architectures have client processes & processes

3 9 Addressing Process - Sockets o process sends/receives messages to/from its socket o socket analogous to door sending process shoves message out door sending process relies on transport infrastructure on other side of door which brings message to socket at receiving process host or process socket TCP with buffers, variables controlled by app developer controlled by OS Internet host or process socket TCP with buffers, variables 10 Application layer protocol defines o Types of messages exchanged, e.g., request, response o Message syntax: what fields in messages & how fields are delineated o Message semantics meaning of information in fields o Rules for when and how processes send & respond to messages Public-domain protocols: o defined in RFCs o allows for interoperability o e.g., HTTP, SMTP Proprietary protocols: o e.g., Skype 11 Transport service that application need Data loss o some apps (e.g., audio) can tolerate some loss o other apps (e.g., file transfer, telnet) require 100% reliable data transfer Timing o some apps (e.g., Internet telephony, interactive games) require low delay to be effective Bandwidth o some apps (e.g., multimedia) require minimum amount of bandwidth to be effective o other apps ( elastic apps ) make use of whatever bandwidth they get 12 Application file transfer Web documents real-time audio/video stored audio/video interactive games instant messaging Transport service requirements of applications Data loss no loss no loss no loss loss-tolerant loss-tolerant loss-tolerant no loss Bandwidth elastic elastic elastic audio: 5kbps-1Mbps video:10kbps-5mbps same as above few kbps up elastic Time Sensitive no no no yes, 100 s msec yes, few secs yes, 100 s msec yes and no

4 13 Internet transport protocols services TCP service: o connection-oriented: setup required between client and processes o reliable transport between sending and receiving process o flow control: sender won t overwhelm receiver o congestion control: throttle sender when network overloaded o does not provide: timing, minimum bandwidth guarantees UDP service: o unreliable data transfer between sending and receiving process o does not provide: connection setup, reliability, flow control, congestion control, timing, or bandwidth guarantee 14 Internet applications : application & transport protocols Application remote terminal access Web file transfer streaming multimedia Internet telephony Application layer protocol SMTP [RFC 2821] Telnet [RFC 854] HTTP [RFC 2616] FTP [RFC 959] proprietary (e.g. RealNetworks) proprietary (e.g., Vonage,Dialpad) Underlying transport protocol TCP TCP TCP TCP TCP or UDP typically UDP Web and HTTP Web & HTTP o Web page consists of objects o Object can be HTML file, JPEG image, Java applet, audio file, o Web page consists of base HTML-file which includes several referenced objects o Each object is addressable by a URL o Example URL: 16 host name path name

5 17 HTTP overview HTTP: hypertext transfer protocol o Web s application layer protocol o client/ model client: browser that requests, receives, displays Web objects : Web sends objects in response to requests o HTTP 1.0: RFC 1945 o HTTP 1.1: RFC 2068 PC running Explorer Mac running Navigator Server running Apache Web 18 HTTP overview (continued) Uses TCP: o client initiates TCP connection (creates socket) to, port 80 o accepts TCP connection from client o HTTP messages (application-layer protocol messages) exchanged between browser (HTTP client) and Web (HTTP ) o TCP connection closed HTTP is stateless o maintains no information about past client requests aside Protocols that maintain state are complex! o past history (state) must be maintained o if /client crashes, their views of state may be inconsistent, must be reconciled 19 HTTP connections Nonpersistent HTTP o At most one object is sent over a TCP connection. o HTTP/1.0 uses nonpersistent HTTP Persistent HTTP o Multiple objects can be sent over single TCP connection between client and. o HTTP/1.1 uses persistent connections in default mode 20 Nonpersistent HTTP Suppose enters URL time 1a. HTTP client initiates TCP connection to HTTP (process) at on port HTTP client sends HTTP request message (containing URL) into TCP connection socket. Message indicates that client wants object somedepartment/home.in dex (contains text, references to 10 jpeg images) 1b. HTTP at host waiting for TCP connection at port 80. accepts connection, notifying client 3. HTTP receives request message, forms response message containing requested object, and sends message into its socket

6 Nonpersistent HTTP (cont.) Non-Persistent HTTP: Response time time 5. HTTP client receives response message containing html file, displays html. Parsing html file, finds 10 referenced jpeg objects 6. Steps 1-5 repeated for each of 10 jpeg objects 4. HTTP closes TCP connection. Definition of RTT: time to send a small packet to travel from client to and back. Response time: o one RTT to initiate TCP connection o one RTT for HTTP request and first few bytes of HTTP response to return o file transmission time total = 2RTT+transmit time initiate TCP connection RTT request file RTT file received time time time to transmit file Persistent HTTP Nonpersistent HTTP issues: o requires 2 RTTs per object o OS overhead for each TCP connection o browsers often open parallel TCP connections to fetch referenced objects Persistent HTTP o leaves connection open after sending response o subsequent HTTP messages between same client/ sent over open connection Persistent without pipelining: o client issues new request only when previous response has been received o one RTT for each referenced object Persistent with pipelining: o default in HTTP/1.1 o client sends requests as soon as it encounters a referenced object o as little as one RTT for all the referenced objects 24 HTTP request message o two types of HTTP messages: request, response o HTTP request message: ASCII (human-readable format) request line (GET, POST, HEAD commands) Carriage return, line feed indicates end of message header lines GET /somedir/page.html HTTP/1.1 Host: User-: Mozilla/4.0 Connection: close Accept-language:fr (extra carriage return, line feed)

7 HTTP request message: general format Uploading form input Post method: o Web page often includes form input o Input is uploaded to in entity body URL method: o Uses GET method o Input is uploaded in URL field of request line: HTTP/1.0 o GET o POST o HEAD Method types asks to leave requested object out of response HTTP/1.1 o GET, POST, HEAD o PUT uploads file in entity body to path specified in URL field o DELETE deletes file specified in the URL field HTTP response message status line (protocol status code status phrase) data, e.g., requested HTML file header lines HTTP/ OK Connection close Date: Thu, 06 Aug :00:15 GMT Server: Apache/1.3.0 (Unix) Last-Modified: Mon, 22 Jun Content-Length: 6821 Content-Type: text/html data data data data data

8 OK HTTP response status codes In first line in ->client response message. A few sample codes: 31 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 404 Not Found requested document not found on this 505 HTTP Version Not Supported client ebay 8734 cookie file ebay 8734 amazon 1678 one week later: ebay 8734 amazon 1678 Cookies: keeping state (cont.) usual http request msg usual http response Set-cookie: 1678 usual http request msg cookie: 1678 usual http response msg usual http request msg cookie: 1678 usual http response msg Amazon creates ID 1678 for create entry cookiespecific action cookiespectific action access access backend database User- state: cookies Many major Web sites use cookies Four components: 1) cookie header line of HTTP response message 2) cookie header line in HTTP request message 3) cookie file kept on s host, managed by s browser 4) back-end database at Web site Cookies (continued) What cookies can bring: o authorization o shopping carts o recommendations o session state (Web ) How to keep state : o protocol endpoints: maintain state at sender/receiver over multiple transactions o cookies: http messages carry state Example: o Susan always access Internet always from PC o visits specific e-commerce site for first time o when initial HTTP requests arrives at site, site creates: unique ID entry in backend database for ID aside Cookies and privacy: o cookies permit sites to learn a lot about you o you may supply name and to sites

9 Web caches (proxy ) Goal: satisfy client request without involving origin o sets browser: Web accesses via cache Proxy o browser sends all HTTP requests to cache client object in cache: cache returns object else cache requests object from origin client, then returns object to client origin origin More about Web caching o cache acts as both client and o typically cache is installed by ISP (university, company, residential ISP) Why Web caching? o reduce response time for client request o reduce traffic on an institution s access link. o Internet dense with caches: enables poor content providers to effectively deliver content (but so does P2P file sharing) Caching example Assumptions o average object size = 100,000 bits o avg. request rate from institution s browsers to origin s = 15/sec o delay from institutional router to any origin and back to router = 2 sec Consequences o utilization on LAN = 15% o utilization on access link = 100% o total delay = Internet delay + access delay + LAN delay = 2 sec + minutes + milliseconds 35 institutional network public Internet 1.5 Mbps access link 10 Mbps LAN origin s institutional cache 36 Caching example (cont) possible solution o increase bandwidth of access link to, say, 10 Mbps consequence o utilization on LAN = 15% o utilization on access link = 15% o Total delay = Internet delay + access delay + LAN delay = 2 sec + msecs + msecs o often a costly upgrade institutional network public Internet 10 Mbps access link 10 Mbps LAN origin s institutional cache

10 Caching example (cont) possible solution: install cache o suppose hit rate is 0.4 consequence o 40% requests will be satisfied almost immediately o 60% requests satisfied by origin o utilization of access link reduced to 60%, resulting in negligible delays (say 10 msec) o total avg delay = Internet delay + access delay + LAN delay =.6*(2.01) secs +.4*milliseconds < 1.4 secs 37 institutional network public Internet 1.5 Mbps access link 10 Mbps LAN origin s institutional cache Conditional GET o Goal: don t send object if cache has up-to-date cached version o cache: specify date of cached copy in HTTP request If-modifiedsince: <date> o : response contains no object if cached copy is up-to-date: HTTP/ Not Modified cache HTTP request msg If-modified-since: <date> HTTP response HTTP/ Not Modified HTTP request msg If-modified-since: <date> HTTP response HTTP/ OK <data> FTP: the file transfer protocol object not modified object modified FTP File Transfer Protocol 40 This image cannot currently be displayed. at host FTP interface FTP client local file system file transfer FTP remote file system o transfer file to/from remote host o client/ model client: side that initiates transfer (either to/from remote) : remote host o ftp: RFC 959 o ftp : port 21

11 41 FTP: separate control, data connections o FTP client contacts FTP at port 21, TCP is transport protocol o client authorized over control connection o client browses remote directory by sending commands over control connection. o when receives file transfer command, opens 2 nd TCP connection (for file) to client o after transferring one file, closes data connection. FTP client TCP control connection port 21 TCP data connection port 20 FTP o opens another TCP data connection to transfer another file. o control connection: out of band o FTP maintains state : current directory, earlier authentication 42 FTP commands, responses Sample commands: o sent as ASCII text over control channel o USER name o PASS password o LIST return list of file in current directory o RETR filename retrieves (gets) file o STOR filename stores (puts) file onto remote host Sample return codes o status code and phrase (as in HTTP) o 331 Username OK, password required o 125 data connection already open; transfer starting o 425 Can t open data connection o 452 Error writing file E - Mail 44 Electronic Mail Three major components: o s o mail s o simple mail transfer protocol: SMTP User Agent o a.k.a. mail reader o composing, editing, reading mail messages o e.g., Eudora, Outlook, elm, Mozilla Thunderbird o outgoing, incoming messages stored on mail SMTP mail SMTP SMTP outgoing message queue mail mailbox

12 45 Electronic Mail: mail s Mail Servers o mailbox contains incoming messages for o message queue of outgoing (to be sent) mail messages o SMTP protocol between mail s to send messages client: sending mail : receiving mail mail SMTP mail SMTP SMTP mail 46 Electronic Mail: SMTP [RFC 2821] o uses TCP to reliably transfer message from client to, port 25 o direct transfer: sending to receiving o three phases of transfer handshaking (greeting) transfer of messages closure o command/response interaction commands: ASCII text response: status code and phrase o messages must be in 7-bit ASCII 47 Scenario: Alice sends message to Bob 1) Alice uses UA to compose message and to bob@someschool.edu 2) Alice s UA sends message to her mail ; message placed in message queue 3) Client side of SMTP opens TCP connection with Bob s mail 1 2 mail 3 4 4) SMTP client sends Alice s message over the TCP connection 5) Bob s mail places the message in Bob s mailbox 6) Bob invokes his to read message mail 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

13 SMTP: final words o SMTP uses persistent connections o SMTP requires message (header & body) to be in 7-bit ASCII o SMTP uses CRLF.CRLF to determine end of message Comparison with HTTP: o HTTP: pull o SMTP: push o both have ASCII command/response interaction, status codes o HTTP: each object encapsulated in its own response msg o SMTP: multiple objects sent in multipart msg Mail message format SMTP: protocol for exchanging msgs RFC 822: standard for text message format: o header lines, e.g., To: From: Subject: different from SMTP commands! o body the message, ASCII characters only header body blank line Message format: multimedia extensions o MIME: multimedia mail extension, RFC 2045, 2056 o additional lines in msg header declare MIME content type (Multipurpose Internet Mail Extensions) MIME version method used to encode data multimedia data type, subtype, parameter declaration encoded data From: alice@crepes.fr To: bob@hamburger.edu Subject: Picture of yummy crepe. MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Type: image/jpeg base64 encoded data base64 encoded data Mail access protocols This image cannot currently be displayed. SMTP SMTP access protocol sender s mail receiver s mail o SMTP: delivery/storage to receiver s o Mail access protocol: retrieval from POP: Post Office Protocol [RFC 1939] authorization ( <-->) and download IMAP: Internet Mail Access Protocol [RFC 1730] more features (more complex) manipulation of stored msgs on HTTP: gmail, Hotmail, Yahoo! Mail, etc.

14 53 POP3 protocol authorization phase o client commands: : declare name pass: password o responses +OK -ERR transaction phase, client: o list: list message numbers o retr: retrieve message by number o dele: delete o quit S: +OK POP3 ready C: bob S: +OK C: pass hungry S: +OK successfully logged on C: list S: S: S:. C: retr 1 S: <message 1 contents> S:. C: dele 1 C: retr 2 S: <message 1 contents> S:. C: dele 2 C: quit S: +OK POP3 signing off 54 POP3 (more) and IMAP More about POP3 o Previous example uses download and delete mode. o Bob cannot re-read if he changes client o Download-and-keep : copies of messages on different clients o POP3 is stateless across sessions IMAP o Keep all messages in one place: the o Allows to organize messages in folders o IMAP keeps state across sessions: names of folders and mappings between message IDs and folder name DNS Domain Name System DNS: Domain Name System Domain Name System: o distributed database implemented in hierarchy of many name s o application-layer protocol host, routers, name s to communicate to resolve names (address/name translation) note: core Internet function, implemented as application-layer protocol complexity at network s edge 56

15 57 DNS DNS services o hostname to IP address translation o host aliasing Canonical, alias names o mail aliasing o load distribution replicated Web s: set of IP addresses for one canonical name Why not centralize DNS? o single point of failure o traffic volume o distant centralized database o maintenance doesn t scale! 58 Root DNS Servers com DNS s org DNS s edu DNS s yahoo.com DNS s Distributed, Hierarchical Database amazon.com DNS s pbs.org DNS s poly.edu umass.edu DNS sdns s Client wants IP for 1 st approx: o client queries a root to find com DNS o client queries com DNS to get amazon.com DNS o client queries amazon.com DNS to get IP address for 59 DNS: Root name s o contacted by local name that can not resolve name o root name : contacts authoritative name if name mapping not known gets mapping returns mapping to local name e NASA Mt View, CA f Internet Software C. Palo Alto, CA (and 36 other locations) b USC-ISI Marina del Rey, CA l ICANN Los Angeles, CA a Verisign, Dulles, VA c Cogent, Herndon, VA (also LA) d U Maryland College Park, MD g US DoD Vienna, VA h ARL Aberdeen, MD j Verisign, ( 21 locations) k RIPE London (also 16 other locations) i Autonomica, Stockholm (plus 28 other locations) m WIDE Tokyo (also Seoul, Paris, SF) 13 root name s worldwide 60 TLD and Authoritative Servers o Top-level domain (TLD) s: responsible for com, org, net, edu, etc, and all toplevel country domains uk, fr, ca, jp. Network Solutions maintains s for com TLD Educause for edu TLD o Authoritative DNS s: organization s DNS s, providing authoritative hostname to IP mappings for organization s s (e.g., Web, mail). can be maintained by organization or service provider

16 61 Local Name Server o does not strictly belong to hierarchy o each ISP (residential ISP, company, university) has one. also called default name o when host makes DNS query, query is sent to its local DNS acts as proxy, forwards query into hierarchy DNS name resolution example recursive query: o puts burden of name resolution on contacted name o heavy load? local DNS dns.poly.edu requesting host cis.poly.edu root DNS authoritative DNS dns.cs.umass.edu TLD DNS 62 DNS name resolution example o Host at cis.poly.edu wants IP address for gaia.cs.umass.edu iterated query: o contacted replies with name of to contact o I don t know this name, but ask this local DNS dns.poly.edu requesting host cis.poly.edu root DNS TLD DNS 6 authoritative DNS dns.cs.umass.edu gaia.cs.umass.edu DNS: caching and updating records o once (any) name learns mapping, it caches mapping cache entries timeout (disappear) after some time TLD s typically cached in local name s Thus root name s not often visited o update/notify mechanisms under design by IETF RFC gaia.cs.umass.edu 63 64

17 65 DNS records DNS: distributed db storing resource records (RR) RR format: (name, value, type, ttl) o Type=A name is hostname value is IP address o Type=NS name is domain (e.g. foo.com) value is hostname of authoritative name for this domain o Type=CNAME name is alias name for some canonical (the real) name is really east.backup2.ibm.com value is canonical name o Type=MX value is name of mail associated with name DNS protocol, messages DNS protocol : query and reply messages, both with same message format msg header o identification: 16 bit # for query, reply to query uses same # o flags: query or reply recursion desired recursion available reply is authoritative DNS protocol, messages Name, type fields for a query RRs in response to query records for authoritative s additional helpful info that may be used 68 Inserting records into DNS o example: new startup Network Utopia o register name networkuptopia.com at DNS registrar (e.g., Network Solutions) provide names, IP addresses of authoritative name (primary and secondary) registrar inserts two RRs into com TLD : (networkutopia.com, dns1.networkutopia.com, NS) (dns1.networkutopia.com, , A) o create authoritative Type A record for Type MX record for networkutopia.com

18 Peer-to-peer 70 P2P file sharing Example o Alice runs P2P client application on her notebook computer o intermittently connects to Internet; gets new IP address for each connection o asks for Hey Jude o application displays other peers that have copy of Hey Jude. o Alice chooses one of the peers, Bob. o file is copied from Bob s PC to Alice s notebook: HTTP o while Alice downloads, other s uploading from Alice. o Alice s peer is both a Web client and a transient Web. All peers are s = highly scalable! P2P: centralized directory P2P: problems with centralized directory original Napster design 1) when peer connects, it informs central : IP address content 2) Alice queries for Hey Jude 3) Alice requests file from Bob centralized directory Bob peers o single point of failure o performance bottleneck o copyright infringement: target of lawsuit is obvious file transfer is decentralized, but locating content is highly centralized Alice 71 72

19 73 75 Query flooding: Gnutella o fully distributed no central o public domain protocol o many Gnutella clients implementing protocol overlay network: graph o edge between peer X and Y if there s a TCP connection o all active peers and edges form overlay net o edge: virtual (not physical) link o given peer typically connected with < 10 overlay neighbors Gnutella: Peer joining 1. joining peer Alice must find another peer in Gnutella network: use list of candidate peers 2. Alice sequentially attempts TCP connections with candidate peers until connection setup with Bob 3. Flooding: Alice sends Ping message to Bob; Bob forwards Ping message to his overlay neighbors (who then forward to their neighbors.) peers receiving Ping message respond to Alice with Pong message 4. Alice receives many Pong messages, and can then setup additional TCP connections Gnutella: protocol Query message sent over existing TCP connections peers forward Query message QueryHit sent over reverse path Scalability: limited scope flooding Query QueryHit Hierarchical Overlay o between centralized index, query flooding approaches o each peer is either a group leader or assigned to a group leader. TCP connection between peer and its group leader. TCP connections between some pairs of group leaders. o group leader tracks content in its children File transfer: HTTP Query QueryHit ordinary peer group-leader peer neighoring relationships in overlay network

20 77 Comparing Client-, P2P architectures Question : How much time distribute file initially at one to N other computers? File, size F Server d N u N u 1 d 1 u s u 2 d2 Network (with abundant bandwidth) u s : upload bandwidth u i : client/peer i upload bandwidth d i : client/peer i download bandwidth 78 Client-: file distribution time o sequentially sends N copies: NF/u s time o client i takes F/d i time to download Time to distribute F to N clients using client/ approach F Server d N u N u 1 d 1 u s u 2 d2 Network (with abundant bandwidth) = d cs = max { NF/u s, F/min(d i ) } i increases linearly in N (for large N) P2P: file distribution time o must send one copy: F/u s time o client i takes F/d i time to download o NF bits must be downloaded (aggregate) F Server d N u N u 1 d 1 u s u 2 d2 Network (with abundant bandwidth) o fastest possible upload rate (assuming all nodes sending file chunks to same peer): u s + Σu i i=1,n Minimum Distribution Time Comparing Client-, P2P architectures P2P Client-Server d P2P = max { F/u s, F/min(d i ), NF/(u s + Σu i) } i i=1,n 80 N

21 P2P Case Study: BitTorrent BitTorrent (1) o P2P file distribution tracker: tracks peers participating in torrent obtain list of peers peer trading chunks torrent: group of peers exchanging chunks of a file o file divided into 256KB chunks. o 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 ) o while downloading, peer uploads chunks to other peers. o peers may come and go o once peer has entire file, it may (selfishly) leave or (altruistically) remain BitTorrent (2) Pulling Chunks o at any given time, different peers have different subsets of file chunks o periodically, a peer (Alice) asks each neighbor for list of chunks that they have. o Alice issues requests for her missing chunks rarest first Sending Chunks: tit-for-tat o Alice sends chunks to four neighbors currently sending her chunks at the highest rate re-evaluate top 4 every 10 secs o every 30 secs: randomly select another peer, starts sending chunks newly chosen peer may join top 4 84 P2P Case study: Skype o P2P (pc-to-pc, pc-tophone, phone-to-pc) Voice-Over-IP (VoIP) application also IM o proprietary applicationlayer protocol (inferred via reverse engineering) o hierarchical overlay Skype login Skype clients (SC) Supernode (SN)

22 Skype: making a call o User starts Skype o SC registers with SN list of bootstrap SNs o SC logs in (authenticate) Skype login o Call: SC contacts SN will callee ID SN contacts other SNs (unknown protocol, maybe flooding) to find addr of callee; returns addr to SC o SC directly contacts callee, overtcp Question? 85