




已閱讀5頁(yè),還剩1頁(yè)未讀, 繼續(xù)免費(fèi)閱讀
版權(quán)說(shuō)明:本文檔由用戶(hù)提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)
文檔簡(jiǎn)介
1英文翻譯:本文出自ComputerNetwork第四版AndrewS.Tanenbaum著NetworkPerformanceMeasurementWhenanetworkperformspoorly,itsusersoftencomplaintothefolksrunningit,demandingimprovements.Toimprovetheperformance,theoperatorsmustfirstdetermineexactlywhatisgoingon.Tofindoutwhatisreallyhappening,theoperatorsmustmakemeasurements.Inthissectionwewilllookatnetworkperformancemeasurements.ThediscussionbelowisbasedontheworkofMogul(1993).Thebasicloopusedtoimprovenetworkperformancecontainsthefollowingsteps:1.Measuretherelevantnetworkparametersandperformance.2.Trytounderstandwhatisgoingon.3.Changeoneparameter.Thesestepsarerepeateduntiltheperformanceisgoodenoughoritisclearthatthelastdropofimprovementhasbeensqueezedout.Measurementscanbemadeinmanywaysandatmanylocations(bothphysicallyandintheprotocolstack).Themostbasickindofmeasurementistostartatimerwhenbeginningsomeactivityandseehowlongthatactivitytakes.Forexample,knowinghowlongittakesforaTPDUtobeacknowledgedisakeymeasurement.Othermeasurementsaremadewithcountersthatrecordhowoftensomeeventhashappened(e.g.,numberoflostTPDUS).Finally,oneisofteninterestedinknowingtheamountofsomething,suchasthenumberofbytesprocessedinacertaintimeinterval.Measuringnetworkperformanceandparametershasmanypotentialpitfalls.Belowwelistafewofthem.Anysystematicattempttomeasurenetworkperformanceshouldbecarefultoavoidthese.MakeSureThattheSampleSizeIsLargeEnoughDonotmeasurethetimetosendoneTPDU,butrepeatthemeasurement,say,onemilliontimesandtaketheaverage.Havingalargesamplewillreducetheuncertaintyinthemeasuredmeanandstandarddeviation.Thisuncertaintycanbecomputedusingstandardstatisticalformulas.MakeSureThattheSamplesAreRepresentative2Ideally,thewholesequenceofonemillionmeasurementsshouldberepeatedatdifferenttimesofthedayandtheweektoseetheeffectofdifferentsystemloadsonthemeasuredquantity.Measurementsofcongestion,forexample,areoflittleuseiftheyaremadeatamomentwhenthereisnocongestion.Sometimestheresultsmaybecounterintuitiveatfirst,suchasheavycongestionat10,11,1,and2oclock,butnocongestionatnoon(whenalltheusersareawayatlunch).BeCarefulWhenUsingaCoarse-GrainedClockComputerclocksworkbyincrementingsomecounteratregularintervals.Forexample,amillisecondtimeradds1toacounterevery1msec.Usingsuchatimertomeasureaneventthattakeslessthan1msecispossible,butrequiressomecare.(Somecomputershavemoreaccurateclocks,ofcourse.)TomeasurethetimetosendaTPDU,forexample,thesystemclock(say,inmilliseconds)shouldbereadoutwhenthetransportlayercodeisenteredandagainwhenitisexited.IfthetrueTPDUsendtimeis300sec,thedifferencebetweenthetworeadingswillbeeither0or1,bothwrong.However,ifthemeasurementisrepeatedonemilliontimesandthetotalofallmeasurementsaddedupanddividedbyonemillion,themeantimewillbeaccuratetobetterthan1sec.BeSureThatNothingUnexpectedIsGoingOnduringYourTestsMakingmeasurementsonauniversitysystemthedaysomemajorlabprojecthastobeturnedinmaygivedifferentresultsthanifmadethenextday.Likewise,ifsomeresearcherhasdecidedtorunavideoconferenceoveryournetworkduringyourtests,youmaygetabiasedresult.Itisbesttoruntestsonanidlesystemandcreatetheentireworkloadyourself.Eventhisapproachhaspitfallsthough.Whileyoumightthinknobodywillbeusingthenetworkat3A.M.,thatmightbepreciselywhentheautomaticbackupprogrambeginscopyingallthediskstotape.Furthermore,theremightbeheavytrafficforyourwonderfulWorldWideWebpagesfromdistanttimezones.CachingCanWreakHavocwithMeasurementsTheobviouswaytomeasurefiletransfertimesistoopenalargefile,readthewholething,closeit,andseehowlongittakes.Thenrepeatthemeasurementmanymoretimestogetagoodaverage.Thetroubleis,thesystemmaycachethefile,soonlythefirstmeasurementactuallyinvolvesnetworktraffic.Therestarejustreadsfromthe3localcache.Theresultsfromsuchameasurementareessentiallyworthless(unlessyouwanttomeasurecacheperformance).Oftenyoucangetaroundcachingbysimplyoverflowingthecache.Forexample,ifthecacheis10MB,thetestloopcouldopen,read,andclosetwo10-MBfilesoneachpass,inanattempttoforcethecachehitrateto0.Still,cautionisadvisedunlessyouareabsolutelysureyouunderstandthecachingalgorithm.Bufferingcanhaveasimilareffect.OnepopularTCP/IPperformanceutilityprogramhasbeenknowntoreportthatUDPcanachieveaperformancesubstantiallyhigherthanthephysicallineallows.Howdoesthisoccur?AcalltoUDPnormallyreturnscontrolassoonasthemessagehasbeenacceptedbythekernelandaddedtothetransmissionqueue.Ifthereissufficientbufferspace,timing1000UDPcallsdoesnotmeanthatallthedatahavebeensent.Mostofthemmaystillbeinthekernel,buttheperformanceutilitythinkstheyhaveallbeentransmitted.UnderstandWhatYouAreMeasuringWhenyoumeasurethetimetoreadaremotefile,yourmeasurementsdependonthenetwork,theoperatingsystemsonboththeclientandserver,theparticularhardwareinterfaceboardsused,theirdrivers,andotherfactors.Ifthemeasurementsaredonecarefully,youwillultimatelydiscoverthefiletransfertimefortheconfigurationyouareusing.Ifyourgoalistotunethisparticularconfiguration,thesemeasurementsarefine.However,ifyouaremakingsimilarmeasurementsonthreedifferentsystemsinordertochoosewhichnetworkinterfaceboardtobuy,yourresultscouldbethrownoffcompletelybythefactthatoneofthenetworkdriversistrulyawfulandisonlygetting10percentoftheperformanceoftheboard.4網(wǎng)絡(luò)性能的測(cè)量當(dāng)一個(gè)網(wǎng)絡(luò)的運(yùn)行效果很差的時(shí)候,它的用戶(hù)通常會(huì)向網(wǎng)絡(luò)運(yùn)行商抱怨并要求提高網(wǎng)絡(luò)的質(zhì)量。為了改善網(wǎng)絡(luò)的性能,網(wǎng)絡(luò)操作人員首先必須確定發(fā)生了什么問(wèn)題。為了找出真正的問(wèn)題所在,操作人員必須進(jìn)行測(cè)量工作。在這一小節(jié)中,我們來(lái)看一看網(wǎng)絡(luò)性能的測(cè)量問(wèn)題。下面的討論以Mogul(1993)的工作為基礎(chǔ)。用來(lái)改善網(wǎng)絡(luò)性能的基本循環(huán)過(guò)程包括以下步驟:(1)測(cè)量有關(guān)的網(wǎng)絡(luò)參數(shù)和性能。(2)試圖理解當(dāng)前的網(wǎng)絡(luò)狀況。(3)改變一個(gè)參數(shù)。這些步驟不斷重復(fù),直到網(wǎng)絡(luò)的性能已經(jīng)足夠好,或者改善性能的全部空間都已經(jīng)被發(fā)掘出來(lái)了。測(cè)量工作可以有許多做法,也可以在許多地點(diǎn)或場(chǎng)所進(jìn)行(既指物理位置,也指協(xié)議棧中的位置)。最基本的一種測(cè)量手段是:在開(kāi)始某一個(gè)動(dòng)作的時(shí)候啟動(dòng)一個(gè)定時(shí)器,然后確定該需要多長(zhǎng)時(shí)間。例如,知道一個(gè)TPDU需要多長(zhǎng)時(shí)間才被確認(rèn)是一個(gè)很關(guān)鍵的測(cè)量指標(biāo)。其他有一些測(cè)量指標(biāo)可以通過(guò)計(jì)數(shù)器來(lái)完成,即記錄某種事件發(fā)生的次數(shù),比如丟失的TPDU的數(shù)量。最后,人們通常對(duì)于某些事物的數(shù)量比較感興趣,比如在特定的時(shí)間間隔內(nèi)所處理的字節(jié)數(shù)。測(cè)量網(wǎng)絡(luò)的性能和參數(shù)有許多潛在的陷阱。以下我們列出其中一部分。任何一種系統(tǒng)化的網(wǎng)絡(luò)性能測(cè)量
溫馨提示
- 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
- 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶(hù)所有。
- 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒(méi)有圖紙預(yù)覽就沒(méi)有圖紙。
- 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
- 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶(hù)上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶(hù)上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
- 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
- 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶(hù)因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。
最新文檔
- 文化創(chuàng)意產(chǎn)業(yè)園區(qū)場(chǎng)地?zé)o償租賃協(xié)議
- 成都企業(yè)股權(quán)變更盡職調(diào)查與代辦合同
- 財(cái)務(wù)部?jī)?nèi)部控制與審計(jì)合同范本
- 常規(guī)量具使用培訓(xùn)
- 氣化爐檢修培訓(xùn)主要內(nèi)容
- 中班組常規(guī)情況匯報(bào)
- 2025年教育事業(yè)統(tǒng)計(jì)培訓(xùn)
- 專(zhuān)科護(hù)士培訓(xùn)計(jì)劃
- 2025新高二(復(fù)習(xí)課)必修下冊(cè)詩(shī)歌知識(shí)梳理(教師版)
- 腫瘤患者春節(jié)節(jié)前健康宣教
- 審計(jì) 第7版 課件 第10章采購(gòu)與付款循環(huán)審計(jì)
- 概率論與數(shù)理統(tǒng)計(jì)(天津理工大學(xué))智慧樹(shù)知到期末考試答案2024年
- 八年級(jí)親子共評(píng)
- 家用冰箱市場(chǎng)調(diào)研報(bào)告
- 國(guó)際財(cái)務(wù)報(bào)告準(zhǔn)則
- 初中數(shù)學(xué)-專(zhuān)項(xiàng)24 圓內(nèi)最大張角米勒角問(wèn)題
- 行政單位酒店住宿合同
- 機(jī)械設(shè)備安裝程序、安裝分類(lèi)、固定方式及安裝新技術(shù)應(yīng)用
- 大樓維修改造工程投標(biāo)方案(完整技術(shù)標(biāo))
- 《建筑施工安全檢查標(biāo)準(zhǔn)》JGJ
- 建筑陶瓷磚檢測(cè)報(bào)告及原始記錄
評(píng)論
0/150
提交評(píng)論