完整版软件工程专业毕业设计外文文献翻译.docx

完整版软件工程专业毕业设计外文文献翻译.docx

《完整版软件工程专业毕业设计外文文献翻译.docx》由会员分享，可在线阅读，更多相关《完整版软件工程专业毕业设计外文文献翻译.docx（11页珍藏版）》请在冰豆网上搜索。

完整版软件工程专业毕业设计外文文献翻译

（

二〇一三年六月

AHISTORICALPERSPECTIVE

Fromtheearliestdaysofcomputers,storingandmanipulatingdataamajorapplicationfocus.Thefirstgeneral-purposeDBMSwasdesignedbyCharlesBachmanatGeneralElectricintheearly1960sandwascalledtheIntegratedDataStore.Itformedthebasisforthenetworkdatamodel,whichwasstandardizedbytheConferenceonDataSystemsLanguages（CODASYL）andstronglyinfluenceddatabasesystemsthroughthe1960s.BachmanwasthefirstrecipientofACM’sTuringAward（thecomputerscienceequivalentofaNobelprize）forworkinthedatabasearea;1973.Inthelate1960s,IBMdevelopedtheInformationManagementSystem（IMS）DBMS,usedeventodayinmanymajorinstallations.IMSformedthebasisforanalternativedatarepresentationframeworkcalledtheAirlinesandIBMaroundthesametime,anditallowedseveralpeopletoaccessthesamedatathroughcomputernetwork.Interestingly,todaythesameSABREsystemisusedtopowerpopularWeb-basedtravelservicessuchasTravelocity!

In1970,EdgarCodd,atIBM’sSanJoseResearchLaboratory,proposedanewdatarepresentationframeworkcalledtherelationaldatamodel.Thisprovedtobeawatershedinthedevelopmentofdatabasesystems:

itsparkedrapiddevelopmentofseveralDBMSsbasedontherelationalmodel,alongwitharichbodyoftheoreticalresultsthatplacedthefieldonafirmfoundation.Coddwonthe1981TuringAwardforacademicdiscipline,andthepopularityofrelationalDBMSschangedthecommerciallandscape.Theirbenefitswerewidelyrecognized,andtheuseofDBMSsformanagingcorporatedatabecamestandardpractice.

Inthe1980s,therelationalmodelconsolidateditspositionasthedominantDBMSparadigm,anddatabasesystemscontinuedtogainwidespreaduse.TheSQLquerylanguageforrelationaldatabases,developedaspartofIBM’sSystemRproject,isnowthestandardquerylanguage.SQLwasstandardizedinthelate1980s,andthecurrentstandard,SQL-92,wasadoptedbytheAmericanNationalStandardsInstitute（ANSI）andInternationalStandardsOrganization（ISO）.Arguably,themostwidelyusedformofconcurrentprogrammingistheconcurrentexecutionofdatabaseprograms（calledtransactions）.Userswriteprogramsasiftheyaretoberunbythemselves,andtheresponsibilityforrunningthemconcurrentlyisgiventotheDBMS.JamesGraywonthe1999TuringawardformanagementinaDBMS.

Inthelate1980sandthe1990s,advancesmadeinmanyareasofdatabasesystems.Considerableresearchcarriedoutintomorepowerfulquerylanguagesandricherdatamodels,andthereabigemphasisonsupportingcomplexanalysisofdatafromallpartsofanenterprise.Severalvendors（e.g.,IBM’sDB2,Oracle8,InformixUDS）developedbynumerousvendorsforcreatingdatawarehouses,consolidatingdatafromseveraldatabases,andforcarryingoutspecializedanalysis.

Aninterestingphenomenonistheemergenceofseveralenterpriseresourceplanning（ERP）andmanagementresourceplanning（MRP）packages,whichaddasubstantiallayerofapplication-orientedfeaturesontopofaDBMS.WidelyusedpackagesincludesystemsfromBaan,Oracle,PeopleSoft,SAP,andSiebel.Thesepackagesidentifyasetofcommontasks（e.g.,inventorymanagement,resourcesplanning,financialanalysis）encounteredbyalargenumberoforganizationsandprovideageneralapplicationlayertocarryoutthesetasks.ThedataisstoredinarelationalDBMS,andtheapplicationlayercanbecustomizedtodifferentcompanies,leadingtolowerIntroductiontoDatabaseSystemsoverallcostsforthecompanies,comparedtothecostofbuildingtheapplicationlayerfromscratch.Mostsignificantly,perhaps,DBMSsofWebsitesstoredtheirdataexclusivelyinoperatingsystemsfiles,theuseofaDBMStostoredatathatisaccessedthroughaWebbrowserisbecomingwidespread.QueriesaregeneratedthroughWeb-accessibleformsandanswersareformattedusingamarkuplanguagesuchasHTML,inordertobeeasilydisplayedinabrowser.AllthedatabasevendorsareaddingfeaturestotheirDBMSaimedatmakingitmoresuitablefordeploymentovertheInternet.Databasemanagementcontinuestogainimportanceasmoreandmoredataisbroughton-line,andmadeevermoreaccessiblethroughcomputernetworking.Todaythefieldisbeingdrivenbyexcitingvisionssuchasmultimediadatabases,interactivevideo,digitallibraries,agenomemappingeffortandNASA’sEarthObservationSystemproject,andthedesireofcompaniestoconsolidatetheirdecision-makingprocessesandminetheirdatarepositoriesforusefulinformationabouttheirbusinesses.Commercially,databasemanage-mentsystemsrepresentoneofthelargestandmostvigorousmarketsegments.Thusthes-tudyofdatabasesystemscouldprovetoberichlyrewardinginmorewaysthanone!

INTRODUCTIONTOPHYSICALDATABASEDESIGN

Likeallotheraspectsofdatabasedesign,physicaldesignmustbeguidedbythenatureofthedataanditsintendeduse.Inparticular,itisimportanttounderstandthetypicalworkloadthatthedatabasemustsupport;theworkloadconsistsofamixofqueriesandupdates.Usersalsorequirementsaboutqueriesorupdatesmustrunorandusers’performancerequirementsarethebasisonwhichanumberofdecisions.

Tocreateagoodphysicaldatabasedesignandtotunethesystemforperformanceinresponsetoevolvinguserrequirements,thedesignerneedstounderstandtheworkingsofaDBMS,especiallytheindexingandqueryprocessingtechniquessupportedbytheDBMS.Ifthedatabaseisexpectedtobeaccessedconcurrentlybymanyusers,orisadistributeddatabase,thetaskbecomesmorecomplicated,andotherfeaturesofaDBMScomeintoplay.

DATABASEWORKLOADS

Thekeytogoodphysicaldesignisarrivingatanaccuratedescriptionoftheexpectedworkload.Aworkloaddescriptionincludesthefollowingelements:

1.Alistofqueriesandtheirfrequencies,asafractionofallqueriesandupdates.

2.Alistofupdatesandtheirfrequencies.

3.Performancegoalsforeachtypeofqueryandupdate.

Foreachqueryintheworkload,wemustidentify:

Whichrelationsareaccessed.

Whichattributesareretained（intheSELECTclause）.

Whichattributesorjoinconditionsexpressedonthem（intheWHEREclause）andtheworkload,wemustidentify:

Whichattributesorjoinconditionsexpressedonthem（intheWHEREclause）and.

ForUPDATEcommands,thefieldsthataremodifiedbytheupdate.

Rememberthatqueriesandupdatestypicallyinvolvesaparticularaccountnumber.Thevaluesoftheseparametersdetermineselectivityofselectionandjoinconditions.

Updatesbenefitfromagoodphysicaldesignandthepresenceofindexes.Ontheotherindexesontheattributesthattheymodify.Thus,whilequeriescanonlybenefitfromthepresenceofanindex,anindexmayeitherspeeduporslowdownagivenupdate.Designersshouldkeepthistrade-offerinmindwhencreatingindexes.

NEEDFORDATABASETUNING

Accurate,detailedworkloadinformationmaybeofthesystem.Consequently,tuningadatabaseafteritdesignedanddeployedisimportant—wemustrefinetheinitialdesigninthelightofactualusagepatternstoobtainthebestpossibleperformance.

Thedistinctionbetweendatabasedesignanddatabasetuningissomewhatarbitrary.

Wecouldconsiderthedesignprocesstobeoveronceaninitialconceptualschemaisdesignedandasetofindexingandclusteringdecisionsismade.Anysubsequentchangestotheconceptualschemaortheindexes,say,wouldthenberegardedasatuningactivity.Alternatively,wecouldconsidersomerefinementoftheconceptualschema（andphysicaldesigndecisionsaffectedbythisrefinement）tobepartofthephysicaldesignprocess.

Wherewedrawthelinebetweendesignandtuningisnotveryimportant.

OVERVIEWOFDATABASETUNING

Aftertheinitialphaseofdatabasedesign,actualuseofthedatabaseprovidesavaluablesourceofdetailedinformationthatcanbeusedtorefinetheinitialdesign.Manyoftheoriginalassumptionsabouttheexpectedworkloadcanbereplacedbyobservedusagepatterns;ingeneral,someoftheinitialworkloadspecificationwillbevalidated,andsomeofitwillturnouttobewrong.Initialguessesaboutthesizeofdatacanbereplacedwithactualstatisticsfromthesystemcatalogs（althoughthisinformationwillkeepchangingasthesystemevolves）.Carefulmonitoringofqueriescanrevealunexpectedproblems;forexample,theoptimizermaynotbeusingsomeindexesasintendedtoproducegoodplans.

Continueddatabasetuningisimportanttogetthebestpossibleperformance.

TUNINGTHECONCEPTUALSCHEMA

Inthecourseofdatabasedesign,wemayrealizethatourcurrentchoiceofrelationschemasdoesnotenableusmeetourperformanceobjectivesforthegivenworkloadwithany（feasible）setofphysicaldesignchoices.Ifso,wemayourconceptualschema（andre-examinephysicaldesigndecisionsthatareaffectedbythechangesthatwemake）.

Wemayrealizethataredesignisnecessaryduringtheinitialdesignprocessorlater,afterthesysteminuseforawhile.Onceadatabasedesignedandpopulatedwithdata,changingtheconceptualschemarequiresasignificanteffortintermsofmappingthecontentsofrelationsthatareaffected.Nonetheless,itmaysometimesbenecessarytorevisetheconceptualschemainlightofexperiencewiththesystem.Wenowconsidertheissuesinvolvedinconceptualschema（re）designfromthepointofviewofperformance.

Severaloptionsmustbeconsideredwhiletuningtheconceptualschema:

Wemaydecidetosettlefora3NFdesigninsteadofaBCNFdesign.

Iftherearetwowaystodecomposeagivenschemainto3NForBCNF,ourchoiceshouldbeguidedbytheworkload.

Sometimeswemightdecide