Unformatted text preview:

A Casefor RedundantArrays of InexpensiveDisks RAID Davtd A Patterson Garth Gibson and Randy H Katz ComputerSaenceD v smn Departmentof ElecmcalEngmeermgandComputerSclencea 571EvansHall Umversity of Cabforma Berkeley CA 94720 partrsl WF kY du Abstract Increasmg performance of CPUs and memorres wrll be squandered lf not matched by a sunrlm peformance ourease m II0 Whde the capactty of Smgle Large Expenstve D T SLED has grown rapuily the performance rmprovement of SLED has been modest Redundant Arrays of Inexpensive Disks RAID based on the magnetic duk technology developed for personal computers offers an attractive alternattve IO SLED promtang onprovements of an or r of mogm e m pctformance rehabdlty power consumption and scalalnlrty Thu paper rntroducesfivclevelsof RAIDS grvmg rheu relative costlpetfotmance and compares RAID to an IBM 3380 and a Fupisu Super Eagle 1 Background Rlsrng CPU and Memory Performance The usersof computersare currently enJoymgunprecedentedgrowth m the speedof computers GordonBell said that between1974and 1984 smgle chip computersimproved m performanceby 40 per year about twice the rate of mmlcomputers Bell 841 In the followmg year B111Joy predictedan evenfastergrowth Joy851 Mamframeand supercomputermanufacturers havmg fficulty keeping pace with the rapId growth predictedby Joy s Law cope by offermg m processors as theu top of the lmeproduct But a fast CPU does not a fast systemmakeGene Amdahlrelated CPU speedto mammemorys12eusmgthis rule Siewmrek821 Each CPU mnstrucaonper second requues one byte of moanmemory If computersystemcostsare not to be dommatedby the cost of memory then Amdahl sconstantsuggeststhat memorychip capacity shouldgrow at the samerate Gordon Moore pr cted that growth rate over 20 years fransuforslclup 2y 1 4 AK predzted by Moore sLaw RAMs have quadrupledm capacity every twotMoom75110threeyeaFIyers861 Recently the rauo of megabytesof mam memoryto MIPS ha9been defti asahha Garcm841 vvlthAmdahl sconstantmeanmgalpha 1 In parl becauseof therapti drop of memoryprices mammemorywe 9have grownfastexthanCPUspeedsandmanymachmesare ppedtoday th alphasof 3 or tigha To mamtam the balance of costs m computer systems secondary storagemustmatchthe advancesm otherpartsof the system A key meas ure of magneuctik technology1sthe growth m the maxnnumnumberof bits that can be stored per squaremch or the bits per mch m a track umes the numberof tracks per mch Called MA D for maxunal area1 density the Fmt Law m Disk Density predicts rank87 MAD lo Year 1971 10 Magnettcdd technologyhasdoubledcapacityandhalvedpnce every three years m hne with the growth rate of semiconductormemory and m practicebetween1967and 1979the dtsk capacityof the averageIBM data processmgsystemmore than kept up with its mammemory Stevens81 Capacity IS not the o rty memory charactensuc that must grow rapidly to mamtam system balance since the speed with which msuuctionsand data are delivered to a CPU also determmesits ulamdte perfarmanceThespeedof mem has tpacefoPtworeasons 1 the mvenuonof caches showmgthat a small buff can be managed automamzallyto containa substanttalfractmnof memoryrefaences 2 and the SRAM technology used to build caches whose speed has lmpmvedattherateof4O tolOO peryear In umtmst to pnmary memory technologres the performance of single large expensive ma8netuzd s SLED hasimproved at a modest rate These mechamcal devu are dommated by the seekand the rotahon delays from 1971 to 1981 the raw seek tune for a high end IBM disk improved by only a factor of two whllt the rocstlon hme did not cbange Harkex811Greaterdensltymeansa lugher transferrate when the mformatmn1sfound andextra headscaneduce theaveaageseektnne but the raw seek hme only unproved at a rate of 7 peryear There 1sno reasontoexpectafasterratemthenearfuture To mamtambalance computersystemshave beenusmgeven larger mam memonesor solid state d s to buffer some of the I O acttvlty This may be a fine solutron for apphcattons whose I O actrvlty has locality of reference and for which volatlltty 1s not an issue but appbcauonsdommatedby a high rate of randommuests for small peces of data suchBStmmact on pmcessmg or by a low numberof requestsfor massive amounts of data such as large simulahons nmnmg on supercomputers arefacmga sermusp mnance hmuatmn 2 The Pendrng I O Crisw What t3 the Impactof lmprovmg the performanceof sOmepiecesof a problem while leavmg othersthe same Amdahl sanswerISnow known asAmdahl sLaw Amdahl67 1 S z 1 n flk Whae S the Pemuswn to copy mthout fee all or w of IS matcnal ISgranted pronded that the COP S zzrcnot made or lstnbuted for dwct commernal advantage the ACM copyright notIce and the tltk of the pubbcatuonand IW da appear and notxe IS en that COPYI K ISby pemtrs on of the Association for Computing Machtnery To COPY otherwIse or to repubbsh requres B fee and or spenfic perm ss o 0 1988ACM 0 89791 268 3 88 OlOP 1 50 effecttvespeedup f fractmnofworkmfastermode and k speedupwhde m fastermode Supposethat somecurrent appbcatmnsspend 10 of thev ume In I G Then when computersare 10X faster accordmgto Bdl Joy m JUSt Overthtte years thenAmdahl sLaw predictsefQcovespeedupwdl be only 5X Whenwe have computerslOOXfaster vmevolutmnof umprcuzessors or by multiprocessors s applrcatlon will be less than 10X faster wastmg90 of thepotenhalspeedup price performanceand rehabduy Our reasoningIS that If there are no advantagesm pnceperformanceor temble d vantages m rehabdlty then there ISIIOneedto explorefurther We chamctenzea transacUon processing workloadto evaluateperformanceof a col Uon of iexpensive d s but remember that such a CollecUonis Just one hardwarecomponentof a completetranacUon processmg system While deslgnmga completeTPS based on these ideas 1senUcmg we will resst that temptaUonm this paper Cabling andpackagmg certamlyan issue m thecostandrehablhty of an array of manymexpenslved s ISalsobeyondthis paper sscope Whde we can lmagme improvementsm software file systemsvia buffcrmg for near term 40 demands we needmnovaUonto avoid an J O crms Boral83 3 A Solution Arrays of Inexpensrve Disks RapIdunprovementsm capacityof largediskshavenot beenthe only targetofd designers smcepersonalcomputershavecreateda marketfor inexpensive magnetic disks These lower cost sks have lower performanceas well as lesscapacity Table I below comparesthe top of the lme IBM 3380 model AK4 mamframedtsk FUJ UM2361A Super Eagle muucomputer disk and the Conner Penpherals CP 3100 personal computerd ChoroctensacS Mainframe Small Computer LJ


View Full Document

Berkeley COMPSCI 152 - A Case for Redundant Arrays of Inexpensive Disks

Documents in this Course
Quiz 5

Quiz 5

9 pages

Memory

Memory

29 pages

Quiz 5

Quiz 5

15 pages

Memory

Memory

29 pages

Memory

Memory

35 pages

Memory

Memory

15 pages

Quiz

Quiz

6 pages

Midterm 1

Midterm 1

20 pages

Quiz

Quiz

12 pages

Memory

Memory

33 pages

Quiz

Quiz

6 pages

Homework

Homework

19 pages

Quiz

Quiz

5 pages

Memory

Memory

15 pages

Load more
Loading Unlocking...
Login

Join to view A Case for Redundant Arrays of Inexpensive Disks and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view A Case for Redundant Arrays of Inexpensive Disks and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?