21v2re4, 420 PM CoMPoSING PReGRAMS. ‘cnaper 4 42.0 Pyhen Steam ‘ueres {£7219 mpbretaion “4p aratel computing 46 Distributed Computing TEXT PROJECTS TUTOR ABOUT 46 Distributed Computing Large-scale data processing applications often coordinate effort among muftiple computers, A distributed computing application is one in which multiple interconnected but independent computers coordinate to perform a joint computation, Different computers are independent in the sense that they do not directly share memory. Instead, they communicate with each other using messages, information transferred from one computer to another over a network. 4.6.1. Messages Messages sent between computers are sequences of bytes. The purpose of a message varies, messages can request data, send data, or instruct another computer to evaluate a procedure cal. In all cases, the sending computer must encode information in a way that the receiving computer ean decode and correctly interpret. To do $0, computers adopt a message protocol that endows meaning to sequences of bytes, A message protocols a set of rules for encoding and interpreting massages. Both the sending and receiving computers must agree on the semantics of # message to enable successful communication. Many message protocols specify that a message conform toa partcular format in which certain bits al fixed positions incicate fixed conditions. Others use special bytes or byte sequences to delimit parts of the message, much as punctuation delimits sub-expressions in the syntax of a programming language. Message protocols are not particular programs or software libraries. Instead, they are rues that can be applied by a variety of programs, even written in different programming languages. As a resull, computers with vastly diferent software systems can participate in the same distributed system, simply by conforming to the message protocols that govern the system. ‘The TCPIP Protocols. On the Internet, messages are transferred from one machine to another Using the Internet Protocol (IP), which specifies how to transfer packets of data among diferent networks to allow global Intemet communication. IP was designed under the assumption that networks are inherently unreliable at any point and dynamic in structure. Moreover, it does not assume that any central tracking or monitoring of communication exists. Each packet contains a header containing the destination IP address, slong with other information. All packets are forwarded ‘throughout the network toward the destination using simple routing rules on a best-effort basi. ‘This design imposes constraints on communication. Packets transferred using modem IP implementations (IPv4 and IPv6) have a maximum size of 85,535 bytes. Larger data values must be split among multiple packets. The IP does nat guarantee that packets will be received inthe same order thal they were sent, Some packets may be lost, and some packets may be transmitted multiple times. ‘The Transmission Control Protocol is an abstraction defined in terms ofthe IP that provides reliable, ordered transmission of arbitrary large byte streams. The protacol provides this guarantee by correctly ordering packets transferred by the IP, emaving duplicates, and requesting retransmission of lost packets, This improved reliability comes atthe expense of latency, the time required to send a ‘message from one point to another. ‘The TCP breaks a stream of data into TCP segments, each of which includes a portion of the data preceded by a header that contains sequence and state information to suppor reliable, ordered transmission of data. Some TCP segments do not include data at all, but instead establish or ‘terminate a connection between two computers Establ'shing a connection between two computers a and 6 proceeds in three steps: 1. sends a request to a port of sto establish a TCP connection, providing a port number to Which lo send the response. 2. esends a response tothe port specified by a and walts for ts response to be acknowledged. 3. asends an acknowledgment response, ver'ying that data can be transferred in both directions. ‘Afr this three-step "handshake", the TCP connection is established, and and s can send data to teach other. Terminating a TCP connection proceeds as a sequence of steps in which both the client and server request and acknowledge the end of the connection 46.2 Client/Server Architecture ‘The clientserver architecture is a way to dispense a service from a central source. A server provides 2 service and multiple clients communicate withthe server to consume that service. In this, nipsilmsew.composingprograms.com/pages/48-distriouted-computing.him| wa21v2re4, 420 PM eT Message Pesing “48. Syremontaton Pals 46 Distributed Computing architecture, clients and servers have different roles, The servers role is to respond to service requests from clients, while a cients roe isto Issue requests and make use of the server's response in order to perform some task, The diagram below illustrates the architecture, ‘The most influential use of the model is the modem World Wide Web. When a web browser displays the contents of a web page, several programs running on independent computers interact using the clientiserver architecture. This section describes the process of requesting a web page in order to ilustrate central ideas in clentserver distributed systems. Roles. Tne web browser application on a Web user's computer has the role ofthe client when requesting a web page. When requesting the content from a domain name on the Internet, such as ‘wurrylimes.com, it must communicate with at least lwo different servers ‘The client frst requests the Internet Protocol (IP) address of the computer located at that name from a Domain Name Server (ONS). A DNS provides the service of mapping domain names to IP addresses, which are numerical identifiers of machines on the Intemet. Python can make such a request direcly using the sacket module. 25> fron socket mport gethosthynane >>> gathostoynane( iw. nytines. com’) 370.149.172.138" ‘The client then requests the contents of the web page from the web server located at that IP adress. ‘The response inthis case is an HTML document that contains headlines and article excerpts of the day's news, as well as expressions that indicate how the web browser client should lay out that contents on the user's sereen. Python can make the two requests required to retrieve this content Using the uritib.request module. >>> fron urllib.nequest dnport urlopen >>> response = urTopentttp:/Ptae.nyt ines. con") read) 353 response 15] b'abocT¥PEntnd>* Upon receiving this response, the browser issues additional requests for images, videos, and other ‘auxiliary components ofthe page. These requests are initiated because the original HTML document contains addresses of additional content and a description of how they embed into the page. ‘An HTTP Request. The Hypertext Transfer Protocol (HTTP) isa protocol implemented using TCP that governs communication for the World Wide Web (WWW). It assumes a client/server architecture betwen a web browser and a web server. HTTP species the format af messages exchanged between browsers and servers. All web browsers use the HTTP format to request pages from a web server, and all web servers use the HTTP format to send back their responses, HTTP requests have several types, the most common of which is a ger request fora specific web page. Acer request specifies a location. For instance, typing the address hetp:/enowikipedta.org/wiki/UC Serkeley inlo a web browser issues an HTTP cer request fo port 80 of the web server at en.wikipedia.org for the contents at location /wiki/Uc_serkeley. “Tho sorver sends back an HTTP response: Server: Apsene/1.3:3.7 (Unix) (Red-Het/Limx) Content-Type: text/her; charset=UTF-8 (On the firstline, the text 260 ox indicates that there were no errors in responding to the request, The subsequent lines of the header give information about the server, the date, and the type of content being sent back, nipsilmsew.composingprograms.com/pages/48-distriouted-computing.him| 2421v2re4, 420 PM 46 Distributed Computing Ifyou have typed in @ wrong web address, or clicked on a broken link, you may have seen a message such as this eror Itmeans that the server sent back an HTTP header that started: HITP/2.2 408 Wot Found ‘The numbers 200 and 404 are HTTP response codes. A fixed set of response codes is a common feature of a message protocol. Designers of protocols attempt to anticipate common messages that will be sent via the protocol and assign fixed codes to reduce transmission size and establish @ common message semantics. In the HTTP protocol, the 200 response code indicates success, while 404 indicates an error that a resource was not found. A varely of other response codes exist in the HTTP 1.1 standard as well Modularity. The concepts of client and server are powerful abstractions. A server provides a service, Possibly to multiple cients simultaneously, and a client consumes thal service, The clients do not need to know the details of how the service is provided, or how the data they are receiving is stored or calculated, and the server does net need to know haw its responses are going to be used. (nthe web, we think of clients and servers as being on different machines, but even systems on a single machine can have clientserver architectures. For example, signals from input devices on a ‘computer need to be generally available to programs running on the computer. The programs are Clients, consuming mouse and keyboard input data. The operating system's device crivers are the servers, taking in physical signals and serving them up as usable input. In addition, the central processing unit (CPU) and the specialized graphical processing unit (GPU) often participate in a Clentserver architecture with the CPU as the client and the GPU as a server of images, [A drawback of client/server systems is thatthe server isa single point of failure. It's the only component withthe abiliy to dispense the service. There can be any number of clients, which are interchangeable and can come and go as necessary. ‘Another drawback of client-server systoms is that computing resources become scarce ifthere are too many clients. Clients increase the demand on the system without contributing any computing 48.3 Peer-to-Peer Systems “The clientserver model is appropriate for service-oriented situations. However, there are other ‘computational goals for which a more equal division of labor is a better choice. The term peer-to-peer is used to describe distributed systems in which labor is divided among all the components of the system, All the computers send and receive data, and they all contribute some processing power and rmomary. AS a distriouted system increases in size, its capacity of computational resources increases. Ina peer-to-peer system, all components of the system contribute some processing power and memory to a distributed computation. Division of labor among all participants isthe identilying characteristic of a peer-to-peer system. This ‘means that peers need to be able to communicate with each other reliably. In order to make sure that messages reach thair intended destinations, pacr-to-peor systems need to have an organized ‘network structure. The components in these systems cooperate to maintain enough information about the locations of ather components to send messages to intended destinations, In some peer-to-pear systems, the job of maintaining the health ofthe network is taken on by a set of specialized components. Such systems are not pure peer-to-peer systems, because they have aifferent types of components that serve different functions, The components that support a peerto- peer network act ke scaffolding: they help the network stay connected, they maintain information about the locations of cifferent computers, and they help newcomers take their place within their neighborhood. “The most common applications of peer-to-peer systems are data transfer and data storage, For data transfer, each computer in the system contributes to send data over the network. Ifthe destination ‘computer isin a particular computer's neighborhood, that computer helps send data along. For data storage, the data set may be too large to fit on any single computer, or too valuable to store on just a single computer. Each computer stores a small portion ofthe data, and there may be multiple copies of the same data spread over diferent computers. When a computer falls, the data that was on it can be restored from other copies and put back when a replacement arrives. ‘Skype, the voice- and video-chat service, is an example of a data transfer application with @ peer-to- peer architecture. When two people on different computers are having a Skype conversation, their communications are transmitted through a peer-to-peer network, This network is composed of other computers running the Skype application. Each computer knows the location of afew other computers in its neighborhoad. A computer helps send a packet to its destination by passing iton a neighbor, which passes it on to some other neighbor, and so on, until the packet reaches its intended destination, Skype is not a pure peer-to-peer system. A scaffolding network of supemodes is nipsilmsew.composingprograms.com/pages/48-distriouted-computing.him| 3421v2re4, 420 PM 4.6 Distributed Computing responsible for logging-in and logging-out users, maintaining information abot computers, and modifying the network structure wiven users enter and exit. Continue: 4.7 Distrbuted Data Processing ‘SengatngPogarstby et Gta, eda etn acu and non ef Can Pagans by Hari Asbon wd Gly Sana sara de Cae nipsulmvew.composingprograms.com/pages/48-distriouted-computing.him| 4