資料分布式計算chapter1slides

1、 Chapter 1 Characterization of Distributed SystemsDistributed Systems: Concepts and DesignDefinition of Distributed Systems zA distributed system is one in which hardware or software components located at networked computers communicate and coordinate their actions only by passing messages. Distribu

2、ted Systems: Concepts and DesignExamples of Distributed Systems zInternetzMobile phone networkszCorporation networkszFactory networkszCampus networkszHome networkszDistributed Systems: Concepts and DesignintranetISPdesktop computer:backbonesatellite linkserver:%network link:%Figure 1.1A typical port

3、ion of the Internet Distributed Systems: Concepts and DesignFigure 1.2A typical intranet the rest of email serverWeb serverDesktopcomputersFile serverrouter/firewallprint and other serversother serversprintLocal areanetworkemail serverthe InternetDistributed Systems: Concepts and DesignIntranet zWha

4、t is Intranet? zA portion of the Internet that is separately administered and has a boundary that can be configured to enforce local security policieszComposed of several LANs linked by backbone connectionszBe connected to the Internet via a routerDistributed Systems: Concepts and DesignFigure 1.3Po

5、rtable and handheld devices in a distributed systemLaptopMobilePrinterCameraInternetHost intranetHome intranetWAP Wireless LANphonegatewayHost siteDistributed Systems: Concepts and DesignFigure 1.4Web servers and web browsersInternetBrowsersWeb servers ProtocolsActivity.htmlhttp:/ /Protocols/Activit

6、y.htmlhttp:/ lsearch?q=kindberghttp:/ /File system of Distributed Systems: Concepts and DesignLayers of ApplicationsOSI 7 Layers ModelData unitLayerFunctionHostlayersData7. ApplicationNetwork process to application. Web application, Client/Server, P2P6. PresentationData representation and encryption

7、. HTML, XML5. SessionInterhost communication. HTTPSegment4. TransportEnd-to-end connections and reliability. TCP, UDPMedialayersPacket3. NetworkPath determination and logical addressing. IPFrame2. Data LinkPhysical addressing (MAC & LLC). Ethernet, ATM, Bit1. PhysicalMedia, signal and binary tra

8、nsmission. wire, wireless, fiber, shared, non-sharedDistributed Systems: Concepts and DesignCharacteristics of Distributed SystemzConcurrencyzconcurrent programs execution share resourcezNo global clockzprograms coordinate actions by exchanging messageszIndependent failureszwhen some systems fail, o

9、thers may not knowDistributed Systems: Concepts and DesignChallengeszHeterogeneityzOpennesszSecurityzScalabilityzFailure handlingzConcurrencyzTransparencyDistributed Systems: Concepts and DesignHeterogeneityzNetworkszEthernet, token ring, etczComputer hardwarezbig endian / little endianzOperating sy

10、stemszdifferent API of Unix and WindowszProgramming languageszdifferent representations for data structureszImplementations from different developerszno application standardsDistributed Systems: Concepts and DesignHeterogeneityzMiddlewarezapplies to a software layer that provides a programming abstr

11、action as well as masking the heterogeneity of the underlying networks, hardware, OSs and programming languages zMobile codezis used to refer to code that can be sent from one computer to another and run at the destinationDistributed Systems: Concepts and DesignOpennesszOpenness of a computer system

12、 zis the characteristic that determines whether the system can be extended and re-implemented in various way. e.g. UnixzOpenness of distributed systems z is determined by the degree to witch new resource sharing services can be added and be made available for use by A variety of client programs. e.g

13、. WebzHow to deal with openness?z key interfaces are published, e.g. RFCDistributed Systems: Concepts and DesignSecurityzConfidentialityzprotection against disclosure to unauthorized individuals, e.g. ACL in Unix File SystemzIntegrityzprotection against alteration or corruption, e.g. checksumzAvaila

14、bility zprotection against interference with the means to access the resources, e.g. Denial of serviceDistributed Systems: Concepts and DesignScalabilityzA system is described as scalablez it will remain effective when there is a significant increase in the number of resources and the number of user

15、szA scalable example system: the Internetzdesign challengeszThe cost of physical resources, e.g., servers support users at most O(n)zThe performance loss, e.g., DNS no worse than O(logn)zPrevent software resources running out, e.g., IP addresszAvoid performance bottlenecks, e.g., partitioning name t

16、able of DNS, cache and replicationDistributed Systems: Concepts and DesignFigure 1.5Computers in the InternetDateComputersWeb servers1979, Dec.18801989, July130,00001999, July56,218,0005,560,8662003, Jan.171,638,29735,424,956Distributed Systems: Concepts and DesignFigure 1.6Computers vs. Web servers

17、 in the InternetDateComputersWeb serversPercentage1993, July1,776,0001300.0081995, July6,642,00023,5000.41997, July19,540,0001,203,09661999, July56,218,0006,598,697122001, July125,888,19731,299,5922542,298,3712003, JulyDistributed Systems: Concepts and DesignFailure handingzDetectingze.g. checksum f

18、or corrupted datazSometimes impossible so suspect, e.g. a remote crashed server in the InternetzMaskingze.g. Retransmit message, standby serverzToleratingze.g. a web browser cannot contact a web serverzRecoveryze.g. Roll backzRedundancyze.g. IP route, replicated name table of DNSDistributed Systems:

19、 Concepts and DesignConcurrencyzCorrectnesszensure the operations on shared resource correct in a concurrent environment e.g. records bids for an auctionzPerformancezEnsure the high performance of concurrent operations Distributed Systems: Concepts and DesignTransparencyzAccess transparencyzusing id

20、entical operations to access local and remote resources, e.g. a graphical user interface with folderszLocation transparencyzresources to be accessed without knowledge of their location, e.g. URLzConcurrency transparencyzseveral processed operate concurrently using shared resources without interferen

21、ce with between themzReplication transparencyzmultiple instances of resources to be used to increase reliability and performance without knowledge of the replicas by users or application programmers, Distributed Systems: Concepts and DesignTransparencyzFailure transparencyzusers and applications to complete their tasks despite the failure of hardware and software components, e.g., emailzMobility transparencyzmovement of resources and clie