文献翻译云计算概述.docx

文献翻译云计算概述.docx

《文献翻译云计算概述.docx》由会员分享，可在线阅读，更多相关《文献翻译云计算概述.docx（14页珍藏版）》请在冰豆网上搜索。

文献翻译云计算概述

本科毕业设计

外文文献及译文

文献、资料题目：

CloudComputing

文献、资料来源：

云计算概述（英文版）

文献、资料发表（出版）日期：

2009年5月

院（部）：

专业：

班级：

姓名：

学号：

指导教师：

翻译日期：

外文文献：

CloudComputing

1.CloudComputingataHigherLevel

Inmanyways,cloudcomputingissimplyametaphorfortheInternet,theincreasingmovementofcomputeanddataresourcesontotheWeb.Butthere’sadifference:

cloudcomputingrepresentsanewtippingpointforthevalueofnetworkcomputing.Itdelivershigherefficiency,massivescalability,andfaster,easiersoftwaredevelopment.It’saboutnewprogrammingmodels,newITinfrastructure,andtheenablingofnewbusinessmodels.

Forthosedevelopersandenterpriseswhowanttoembracecloudcomputing,Sunisdevelopingcriticaltechnologiestodeliverenterprisescaleandsystemicqualitiestothisnewparadigm:

（1）Interoperability—whilemostcurrentcloudsofferclosedplatformsandvendorlock-in,developersclamorforinteroperability.Sun’sopen-sourceproductstrategyandJava™principlesarefocusedonprovidinginteroperabilityforlarge-scalecomputingresources.Thinkoftheexistingcloud“islands”mergingintoanew,interoperable“Intercloud”whereapplicationscanbemovedtoandoperateacrossmultipleplatforms.

（2）High-densityhorizontalcomputing—Sunispioneeringhigh-power-densitycompute-nodearchitecturesandextreme-scaleInfinibandfabricsaspartofourtop-tierHPCdeployments.Thishigh-densitytechnologyisbeingincorporatedintoourlarge-scaleclouddesigns.

（3）Datainthecloud—Morethanjustcomputeutilities,cloudcomputingisincreasinglyaboutpetascaledata.Sun’sOpenStorageproductsofferhybriddataserverswithunprecedentedefficiencyandperformancefortheemergingdata-intensivecomputingapplicationsthatwillbecomeakeypartofthecloud.

Thesetechnologybetsarefocusedondrivingmoreefficientlarge-scaleclouddeploymentsthatcanprovidetheinfrastructurefornext-generationbusinessopportunities:

socialnetworks,algorithmictrading,continuousriskanalysis,andsoon.

2.WhyCloudComputing?

（1）Clouds:

MuchMoreThanCheapComputing

CloudcomputingbringsanewlevelofefficiencyandeconomytodeliveringITresourcesondemand—andintheprocessitopensupnewbusinessmodelsandmarketopportunities.

Whilemanypeoplethinkofcurrentcloudcomputingofferingsaspurely“paybythedrink”computeplatforms,they’rereallyaconvergenceoftwomajorinterdependentITtrends:

ITEfficiency—MinimizecostswherecompaniesareconvertingtheirITcostsfromcapitalexpensestooperatingexpensesthroughtechnologiessuchasvirtualization.Cloudcomputingbeginsasawaytoimproveinfrastructureresourcedeploymentandutilization,butfullyexploitingthisinfrastructureeventuallyleadstoanewapplicationdevelopmentmodel.

BusinessAgility—MaximizereturnusingITasacompetitiveweaponthroughrapidtimetomarket,integratedapplicationstacks,instantmachineimagedeployment,andpetascaleparallelprogramming.Cloudcomputingisembracedasacriticalwaytorevolutionizetimetoservice.Butinevitablytheseservicesmustbebuiltonequallyinnovativerapid-deployment-infrastructuremodels.

Tobesure,thesetrendshaveexistedintheITindustryforyears.However,therecentemergenceofmassivenetworkbandwidthandvirtualizationtechnologieshasenabledthistransformationtoanewservices-orientedinfrastructure.

CloudcomputingenablesITorganizationstoincreasehardwareutilizationratesdramatically,andtoscaleuptomassivecapacitiesinaninstant—withoutconstantlyhavingtoinvestinnewinfrastructure,trainnewpersonnel,orlicensenewsoftware.Italsocreatesnewopportunitiestobuildabetterbreedofnetworkservices,inlesstime,forlessmoney.

ITEfficiencyonaWholeNewScale

Cloudcomputingisallaboutefficiency.ItprovidesawaytodeployandaccesseverythingfromsinglesystemstohugeamountsofITresources—ondemand,inrealtime,atanaffordablecost.Itmakeshigh-performancecomputeandhigh-capacitystorageavailabletoanyonewithacreditcard.Andsincethebestcloudstrategiesbuildonconceptsandtoolsthatdevelopersalreadyknow,cloudsalsohavethepotentialtoredefinetherelationshipbetweeninformationtechnologyandthedevelopersandbusinessunitsthatdependonit.

Reducecapitalexpenditures—CloudcomputingmakesitpossibleforcompaniestoconvertITcostsfromcapitalexpensetooperatingexpensethroughtechnologiessuchasvirtualization.

Cutthecostofrunningadatacenter—Cloudcomputingimprovesinfrastructureutilizationratesandstreamlinesresourcemanagement.Forexample,cloudsallowforself-serviceprovisioningthroughAPIs,bringingahigherlevelofautomationtothedatacenterandreducingmanagementcosts.

Eliminateoverprovisioning—Cloudcomputingprovidesscalingondemand,which,whencombinedwithutilitypricing,removestheneedtooverprovisiontomeetdemand.Withcloudcomputing,companiescanscaleuptomassivecapacitiesinaninstant.

Forthosewhothinkcloudcomputingisjustfluff,takeacloserlookatthecloudofferingsthatarealreadyavailable.MajorInternetprovidersA,Google,andothersareleveragingtheirinfrastructureinvestmentsand“sharing”theirlarge-scaleeconomics.AlreadythebandwidthusedbyAmazonWebServices（AWS）exceedsthatassociatedwiththeircoree-tailingservices.Forward-lookingenterprisesofalltypes—fromWeb2.0startupstoglobalenterprises—areembracingcloudcomputingtoreduceinfrastructurecosts.

Faster,MoreFlexibleProgramming

Cloudcomputingisn’tonlyabouthardware—it’salsoaprogrammingrevolution.Agile,easy-to-access,lightweightWebprotocols—coupledwithpervasivehorizontallyscaledarchitecture—canacceleratedevelopmentcyclesandtimetomarketwithnewapplicationsandservices.Newbusinessfunctionsarenowjustascriptaway.

Acceleratedcycles—Thecloudcomputingmodelprovidesafaster,moreefficientwaytodevelopthenewgenerationofapplicationsandservices.Fasterdevelopmentandtestingcyclesmeansbusinessescanaccomplishinhourswhatusedtotakedays,weeks,ormonths.

Increaseagility—Cloudcomputingaccommodateschangelikenoothermodel.Forexample,AnimotoProductions,makersofamashuptoolthatcreatesvideofromimagesandmusic,usedcloudcomputingtoscaleupfrom50serversto3,500injustthreedays.Cloudcomputingcanalsoprovideawiderselectionofmorelightweightandagiledevelopmenttools,simplifyingandspeedingupthedevelopmentprocess.

Theimmediateimpactwillbeunprecedentedflexibilityinservicecreationandaccelerateddevelopmentcycles.Butatthesametime,developmentflexibilitycouldbecomeconstrainedbyAPIsifthey’renottrulyopen.Cloudcomputingcanusherinaneweraofproductivityfordevelopersiftheybuildonplatformsthataredesignedtobefederatedratherthancentralized.Butthere’samajorshiftunderwayinprogrammingcultureandthelanguagesthatwillbeusedinclouds.

Today,theintegrated,optimized,open-sourceApache,MySQL,PHP/Perl/Python（AMP）stackisthepreferredplatformforbuildinganddeployingnewWebapplicationsandservices.Cloudcomputingwillbethecatalystfortheadoptionofanevennewerstackofmorelightweight,agiletoolssuchaslighttpd,anopen-sourceWebserver;Hadoop,thefreeJavasoftwareframeworkthatsupportsdata-intensivedistributedapplications;andMogileFS,afilesystemthatenableshorizontalscalingofstorageacrossanynumberofmachines.

（2）CompellingNewOpportunities:

TheCloudEcosystem

Butcloudcomputingisn’tjustaboutaproliferationofXenimagestacksonarestrictedhandfulofinfrastructureproviders.It’salsoaboutanemergingecosystemofcomplementaryservicesthatprovidecomputingresourcessuchason-rampsforcloudabstraction,professionalservicestohelpindeployment,specializedapplicationcomponentssuchasdistributeddatabases,andvirtualprivatedatacentersfortheentirerangeofITprovidersandconsumers.

Theseservicesspantherangeofcustomerrequirements,fromindividualdevelopersandsmallstartupstolargeenterprises.Andtheycontinuetoexpandthelevelsofvirtualization,akeyarchitecturalcomponentofthecloudthatoffersever-higherabstractionsofunderlyingservices.

（3）HowDidCloudComputingStart?

Atabasiclevel,cloudcomputingissimplyameansofdeliveringITresourcesasservices.AlmostallITresourcescanbedeliveredasacloudservice:

applications,computepower,storagecapacity,networking,programmingtools,evencommunicationsservicesandcollaborationtools.

Cloudcomputingbeganaslarge-scaleInternetserviceproviderssuchasGoogle,Amazon,andothersbuiltouttheirinfrastructure.Anarchitectureemerged:

massivelyscaled,horizontallydistributedsystemresources,abstractedasvirtualITservicesandmanagedascontinuouslyconfigured,pooledresources.ThisarchitecturalmodelwasimmortalizedbyGeorgeGilderinhisOctober2006Wiredmagazinearticletitled“TheInformationFactories.”TheserverfarmsGilderwroteaboutwerearchitecturallysimilartogridcomputing,butwheregridsareusedforlooselycoupled,technicalcomputingapplications,thisnewcloudmodelwasbeingappliedtoInternetservices.

Bothcloudsandgridsarebuilttoscalehorizontallyveryefficiently.Botharebuilttowithstandfailuresofindividualelementsornodes.Botharechargedonaper-usebasis.Butwhilegridstypicallyprocessbatchjobs,withadefinedstartandendpoint,cloudservicescanbecontinuous.What’smore,cloudsexpandthetypesofresourcesavailable—filestorage,databases,andWebservices—andextendtheapplicabilityto