Chapter 11: Storage and File Structure

صفحه 1:
Chopter 00: Gtorage wad Pile Grwiure 

صفحه 2:
عسنصطت) ۳۲) لمه مت :10 سومان Overview of ‏موه موه‎ Decks ‏ساسم(‎ Disks ROW] ما رو prep ‏سوت‎ مس ۳ Orcerézaioa of Records ia Plies Octe-Dirtrmary Corece له اسان و سس مومس 4 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ae ©Sbervehnts, Cork ced Cnakershe 

صفحه 3:
+ ‏سین‎ oP Physiod Craw Orda Bl Gpeed wih whik chia xn be uerevsed Opt per vit oP char B Rekbhy © choc bree oa power Pakire or syste orcck ۶ ‏موم‎ Pakwe of the storage device 13 ‏مون‎ stanne fib! © ‏ادا‎ storage! bees oouiects wheat power ts swicked oP ۰ ‏اجه وی‎ ‏عون(‎ persist evew wheo power ts switched ‏باه‎ ‎sepoedary ood tertary storage, oe well os butter-bocked‏ ولا ‎Me wokrwewory.‏ ‎Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ae ©Sbervehnts, Cork ced Cnakershe 

صفحه 4:
Physird Gorage Deda ۲ ‏عون‎ - Postest und west costy Por of storage; ‏بط لیم امس‎ copier systec hardier. یی بو ا" 7 2 امس ) زط سس و 1002 تا 102) مه م۳ و (علجسد © ‏و بت رل‎ (or too expewsive) to store he ecire ‏بل‎ ‎١ ‏حوس‎ of up tra Pew Birches widely ‏تاه ای‎ > Coparties hove yor up ced perbyte costs have deorecsed steudfy ord reply (ahh Porter of © every © i 9 pears) ۴ ‏و تن عون - سلاو‎ wewory ore usudly Ist Po power Poke or systew ‏و‎ vos. 4 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ae ©Sbervehnts, Cork ced Cnakershe 

صفحه 5:
+ ‏سس‎ Oorag Deda (Oru) © Chek wom © 055 ‏«وصتخصد‎ power Poke © Qo cos be writes oo boron ody ore, but location coc be erceed ocd vores to cpa ١ Cant support ody 3 keoted carober (DK — (0) of wrtelerase aes. © Groen oP rewory hee i be doce ‏هه و‎ back of ‏ی‎ ‎rank ore reunhy ur Peet we wis weary ut wrtes ore slow (Pew wieroseroads), erase ts sbwer Cost per vat oP storage rouchiy sicker tp wie weeny Ode) Word tn ewobedded devices suck os chet camera 45 ape oF BEPROO (Cloiredly ‏موم لس‎ Read Oct: Dewory) سا0 لح 0 لا سواه 1 6 ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 6:
+ ‏سس‎ Oorag Deda (Oru) شین ۲ سسوم ال له لس موه ‎Oats stored va‏ © © Oriavary wedier Por the loapiers storage of data; ypicdly stores etre ‏بل‎ © Dota wet be woved Prow disk to wait wewory Por uovess, ued varied bac Por store ١ Dich sewer coceee hoa wo ‏موه‎ (core va this kter) ۰ ‏مین‎ — posable to read chia oa disk ‏ماو ی له زو و‎ fare ard deks us Poppy coke Depceies rome uy ty ray POO ۵۵ ‏تاه‎ ‎» Dyck larwer copay end costlbye thor cots recor /Pkeok ecm و © ‎wih eckorkw koproveweds (Paster of‏ رو له رام مس ‎pews)‏ © مد 5 © Gunes power Pokres wad sysiew ‏اس‎ ‎0 disk Pakire can destroy data, but ie rare سا0 لح 0 لا سواه 1 هه ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 7:
+ ‏سس‎ Oorag Deda (Oru) Bl Opted spar © exe unk, dott i ‏موه مس رل ام‎ tok utero hase © CORO (FO OO) cad DOD (P.P WP CO) covet popu Pore © Orie, reatoay (DORO) opted debe wed Por achid stone (CO- &, DOOR, DOO+R) © ‏ان(‎ wrte versizes do wakble (OO-RO, DOO-RO, DOOHRO, exert OOO-REO) © Reads oad writes ore slower too wil: wocaetio dk بوعشل تجاه ,حاط جاتاسومم اه رای ‎wis large‏ رم و بط ول یه ساره عصل تن بای لها وله ۳ منطو ولمم ‎dak‏ اه وس با 1 Ovweyte- O* Btn, Ours, OOOO. ar ©Sbervehnts, Cork ced Cnakershe 

صفحه 8:
+ ‏سس‎ Oorag Deda (Oru) ا ‎ervchiile, sed primary Por bockup (io recover Prox dk Pohure), ocd‏ © ‎Por archival dota‏ ۶ ‏ای - سمل هو‎ slower trom dob ‏رده رجا بسن‎ )00۵ OOO GO tapes wake) 8 tape cam be ‏نايك صا لصم‎ = strane costs wk ‏بحاصل ما واه‎ bidtes oF expe © Rape Nkoboxes wokkble Por stork cxssive coo nis oP cht ۱ ‏لیا‎ oP torches (ercbyie = (0° bytes) to even pete (( pete = 10 byez) Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «oe ©Sbervehnts, Cork ced Cnakershe 

صفحه 9:
ain memory sh memory 1 agnetic disk optical disk magnetic tape: ۳ 1 Ovweyte- O* Btn, Ours, OOOO. 

صفحه 10:
+ Garap Deracly (lod) ‎pekowy sera? Posteot wed bt uchile (cocke, wok cower).‏ لا ‎wert level nt hierarchy, wovcktie, woderdely Post acess‏ تس جوت موی ۴ ‎Cd ‏مه وی لسن سل‎ © Cy, Peck ‏صحاطك و رو‎ ‎8 ertey sprap? best vel io herochy, wcruckile, slow ‏وه‎ koe © bo odled oP Mie storage © ‏اوه روا مومت بر‎ sire ‎1 Ovweyte- O* Btn, Ours, OOOO. ann ©Sbervehnts, Cork ced Cnakershe 

صفحه 11:
باطاتصعففة حصي 000۵: ‏عبت اف لت و ده سا امس لت ساره < و0‎ Ez Deedee Ort Overy - O* Ctorn, Ours, OOOO. aca ©Sbervehnts, Cork ced Cnakershe 

صفحه 12:
Quays Disks Reaburte howd © ‏حيست برص لجس‎ tthe phiter surPease (coo truckion t) © Reb or mies ‏لس هون‎ Porto, ‎track‏ موی ند الق ان ‎oP‏ سای ‎© Over 00۷0006 racks: per phater va ypied ‏سا او‎ ‏,وه با الط و ‎(Bark rok‏ ‎9 0 ‏نمی سوه باب مه‎ chats tral eam ber reed or Lerten, ‏متا 946 رش سحاد سس © ‎۶ ‏مسر سوه لو‎ rack! GOO (ox ker tracks) ty (DDO (oa eter rock) Do renlhurte ‏ممصت د‎ ‎© ‏حاص‎ or owing posts head right racks ‎© per opie out); cata by readlrtion oy sevior pares urder heed Ahexnideh wererbben ‎© cnthple del phitery oa tice yptade (to G verily) ‏و موی و و لت ان ‎poe head‏ 8 ‎9 alle platens ‎ ‎Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ase ©Sbervehnts, Cork ced Cnakershe 

صفحه 13:
+ ‏مه‎ 0 (Ova) © Gorter ‏و مج‎ were ‏نامه‎ to head-orushes © GePove oP curler ‏ارت مه سای لا علاط مس‎ would ‏سپس‎ oo head arash ood daczace oll dota cr cts © Cured yecerdivg disks ore less susceptible to suck deustrous Puhures, ‏رو وه الط ماه‎ yet corrupted: سا له موه وه با مسا اما یت ‎Ok‏ © لها © coven hightevel cowwaeds ty ‏و هط و ام‎ 8 tities untivas suck os woviey the dish are te the right track cred actudhy recdiory pr writer the cake © Computes oad utaches chevhsuws ty rack senior ty verPy thot dota te read back ‏باس‎ 2B data te corrupted, wih very high probably stored chechsue wea wots checks © ‏مج‎ sucess Pl varios by ready back sector Per vrticrs tt © ‏مرخب‎ rewoppiay of bud seviors Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ase ©Sbervehnts, Cork ced Cnakershe 

صفحه 14:
system bus disk controller ۲9 9 9 9 ای و جوا موه وی و با موی سا ای( 191 ۶ ‏وه لا رجام) رطس( مسیون‎ remap) Pied carried out by seat nt deers) aches: bend om erent Bik iterPace otxntards Pole © PDD (OO ‏له‎ rene oF skeemkarcs © GON (Gered BPO) © ever vars of ruck standard (dP Perrd speeds on cocbtiers) Deedee Orie ‏مسف “© - عابس‎ Ours, OOOO. ane ‏لاس09‎ ۱ ‏و‎ standards 8 

صفحه 15:
+ PaPorwous Onanies ch Dito Bl ‏یوق‎ toe — he neve i tchew Prow whee a read or wre request by ‏ميل مایت نا اس‎ irnePer becker. Orcs of © Geek teow — oe I tohew to repention ther are over he omen irark. © Dera geeks te ty UE the worst oxen seeks tere - Drak be (1/9 Pd rocky hed he sexe oxnvber of ‏ما سید سب لو موه‎ tke ‏و‎ stat ox stop are wovec ١ (0 wllseernbs ow ‏ور‎ doko ۶ ٩9م ‏روا‎ — thee i thes Por he sector io be once to appear ener the head. ‏رها وم‎ VE oP te wernt coor hice). ١ 4: ‏د‎ 00 volloernnbs om ytd toby (GPOO & (SOOO rp.co.) Bl Octrtexe Per rate — he rate ot hick dt oun be retreved Prax or stored to the dk, © 2S & WOO WO per sewed wax rate, buwer Por keer chy © Dade oho way shure a eirider, ‏جماتصمت نجل نقح برت‎ von hunny ‏مج سای‎ ١ Ox. OOS: 95 OObev, COTO: (SO Olver, Dira FEO ۵061: 0 0۶ < ‏اسان سا‎ )306(: CSO OV/s 1 Ovweyte- O* Btn, Ours, OOOO. ase ©Sbervehnts, Cork ced Cnakershe 

صفحه 16:
+ 602552525250552 0 Bl Dean tere t Poker (DT) — he werene the he dob tr expected iy net pone wake cy ‏.سانا‎ © ‏حجر‎ 9 te S pears © ‏الا‎ oP ‏و ای(‎ cew dike ie quie low, correspoudiay to “heorricd ODD" be GOO,OOO te 1,800,000 hows Por uve chests * Exy., co OPNE cf 6,E0O0,000 hows Por a vew disk weuus thot ‏مسق‎ (OOO retiively cew disks, vo oo were vor wil Pal every (200 hare © OPVE devrewes us disk oer Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ase ©Sbervehnts, Cork ced Cnakershe 

صفحه 17:
+ Optottiva oP Oek-Biok ‏وه‎ © @bck — ‏وه موی و‎ oP sectors Proc 3 staf track © dots exePerred between disk ood wait wewory ta blocks © see rene Prow GUC bytes ty severd hivbyies * Goodker blocks: wore traasPers Pro dts * Larger blocks? wore spare wrested due te portly Piled blacks ١ OD ypiod block sizes today rece Proc ۴ ‏روط 16 و‎ ۲ ‏وه بط مه جوا مش الیو وی‎ ib hacks ‏جلف فا ود‎ ‏لس سا موه و‎ © ‏مرول رو‎ : re doh ane fa oe directo (Prow outer i Keer treks oF Ue versu), ‏پیج‎ wed request mira devia, lo wore renest ni ho drevim, feu reverse devi onl repeal Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 18:
+ Opteotzcaoa oP Dek Book Bovess (Ova) تا رم ‎koe by‏ مه را ‎Be orpctratoa — opikoize‏ ات ال توا وا ‎correspond‏ 8 ‏من مت لوا سوه وق‎ he sexve ‏لاه وه و‎ © Chee ‏لب رم‎ over oe ۱ Gig. P dota is ‏لاله لصو‎ Proc the Pe ” Or Pree blocks va disk ore ‏له له‎ oeuly orected Pe be ite blocks sruttered over the dest ١ Gequectal ascess ty Progeoected Pie resale ta tooreused disk arc ‏سوه‎ © Gowe systews hove utes to ‏,روط با اعد‎ ta order to speed vp Pie ures Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ase ©Sbervehnts, Cork ced Cnakershe 

صفحه 19:
+ Opteotzcaoa oP Dek Book Bovess (Ova) B Dowohale write bP Pers opeed up deb unites by wate bloke i a ewervoktie ROD bP Per ‏امد‎ © Dovevoake ۹000: ‏ای نان 0100 ری لاتم ترا‎ © Crea P power Pale, te dota ty uP aed ul be write to debs ube power retarcey مسا تاسسوم و تسم جات بت نع ‎Coxroler thea wurtes to doh wherever be dob‏ © مه موی ما مسر مت مج مت انا لموته رات تما بقل موم فا من له( ‎without uzeatery Por deta ty ber writes cts‏ ٩ ‏معدي ترجف ال سفن رد لو بسا نوت هط‎ ‏سم باب و - با باق‎ write ‏سای اسان ان رما هیوست ه‎ 9 ‏مامت سا رامیت لوط‎ RDO > Orte wy bu dok b ‏ند مه سوه اه بر‎ ore rewntresd > Do werd Por opretd harchuare (DO-ROO) Bie sete breil) reorder wren ‏مسجم سلجي بممصومت بحاصل م‎ © loxerwchor Plo syetrere Ure skis i 7 © Reorder wikis jour rok oP ‏مرت‎ of Ply spot dora Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ase ©Sbervehnts, Cork ced Cnakershe 

صفحه 20:
+ Row ۲ ROW: Reduerdoat Orraye of ‏)سبط‎ ۱ techotques hot wormage a forge oueobers oP dishes, provickrg ot views oP a siento de oP ۱ ack papaniy od high speed by ustay supe desks tt parcel, ccd * ick relobiliy by stortag cata redvedoaly, 90 thot cok coc be recovered eved iP va ists Peaks © Phe chowe thot sewer dsb out oP a set oP OD disks will Pal is uch higher thos the: ‏ساره تمه و ما وه‎ disk wil Pal. ۶ ‏ما 10۵,0۵۵ ۵۳ ۳۱۸ لت ات ,تعاط 0100 لب موه ربق‎ (cerrom. (pews), ud have a yt DPD of (DOO hows (approx. 0 chs) امه ها نت لیس بقل لاه و ولا ‎Por wien‏ و © ‎oP dhe:‏ BE Orkid) o oosteProie dennive b hw, expeusie dks © 1115 ROT ort ‏"تسوس “سا لحصاد‎ © Deda RO1Ds oe wed Por the higher rekebly ond bork. » Dhe “1” ts interpreted us tdepecdect 8 1 Ovweyte- O* Btn, Ours, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 21:
+ ‘koprovewest oP Rehabhiy via Reduadauy BH Qeduedawy — sore extra iPorwoica trot cad be used to rebudd iePorcaticg ost toa disk Poke B Cx, Orrortag (or ) © Dupkecte every deck. bodied deck cocsiis oP we physind ‏بحاصف‎ ‎© Coens wnte & carted out va bok debe » Reade con tok place Prow ether dk © AP ove desk iva poir Pols, cota sill avakdble fa the oer ١ ‏ا مق‎ wank porwr ody Bo dk Pals, ond ts wirror cok de Pale bePore the syste i repured — Probably oP coobied eved i very sor « ‏تمسق‎ ۳ depeuntedt Palure wades suck os Pre or bulky ‏وتام‎ or ekepiicd power sues Bl Dew koe ty dott bow ‏باه( نا سم مس لوط‎ ‏و‎ ‎© Gx OPE 40۵,۵۵۵ hers, weer tere to repo oF UD ‏مد‎ eer eoeue fare thts bor oP SDO*UD? hows (or SP, DOO peers) Por ‏ب‎ ‏مت روط تاه مر او‎ Peakare crs) Ez Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. eo ©Sbervehnts, Cork ced Cnakershe 

صفحه 22:
+ 4| via (Parcels وا ون ماو ‎of‏ عون مت تا ۴ احجط و و موه لد ول سل لا ‎٩.‏ ‎then,‏ مور تخیر با سوواط 3:۳ اس سای سوه مد رم برجا جف ای سر لا revel sitptag — opt he bis of rack byte carves uliple doh © Aaa arr of eit doko, arte bi fof eas byte to disk ‎red du ot exit ves he rae oP u ode doh.‏ مت مت انس ‎© Ou sechlawess ie Worse thon Por a skier dots ١ Bt evel ‏موه‎ te wot wed cack ey wore ‎Bl Whole burl striptag — wth ‏حاصف ب‎ block oP a Ple «pes to dk (tooo «) + 0 © Reese Por dPPered books man nc pardel P he books reside ot ‎Perec dishes ‎© @ request Por aint sequewe of books ous uilze ol coke ts parcel ‏سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 23:
+ ROW Loved ۳ Gchewes provide ‏جر امس موه ی وی بو اه واه ول‎ pony bie © OF Foret ROW overtones, or ROD kevets, have dhPericn ost, perPonmue ool reubiay charter BROW Level O: ‏مدل سس موه و‎ © Osed ‏وتو و‎ applic where det bot et orca. BROW Level: Oirored debe wth beck ‏موه‎ ‎© OP Pere best write perPonmane. © Copter Por oppledions suck oF store fey Ales ‏روط و‎ ۱ ۹ (a) RAID 0: nonredundant striping feed (eee Goo becca bubs (b) RAID 1: mirrored disks Ez موم ‎Octo, Ours, OOOO.‏ 0۰ و6 6 یه 

صفحه 24:
+ ROW Level (Ovu.) © ROW bed ©: Dewor-Site Grro~Oorreviter Codes (CO) wth bt striptrg. BROW Level 9: Or oierteaed Poy © cake pany bi i ean Por error ‏ام مه‎ het deterion, stare we kom hick desk beer Poked * Oke voriter ‏روص رل‎ poriy bis cust us be oowpuied ood writes toa porty bi cists: » Do recover chia ina ckemaned deh, compu KOR of bits Brow riker debe (rack pry bt dish) bouts (c) RAID 2: memory-style error-correcting codes (a) RAID 3: bit-interleaved parity Ez ‎Oe BObervchnts, Cork wed Crakershe‏ 7 ۵7 بط 0۰ - ات6 6 تیه ‎ 

صفحه 25:
+ ROW Level (Ovu.) BROW Lael 9 (Cru) ۶ Poster dtc ire Per thon wil ot steer disk, but Pewer VOs per seooed soe every toh ker to ‏سپ‎ in every VO- © Caheanves Level © (provides ol te bouts, ot knver cost). BROW Led P: Oho terbaed Pony; wees borleeyel stripy, ood heeps 3 ‏نما بقع‎ vou separa: dek Por ‏سلجم سمت‎ book Prow D uker doko. © Okea wartiec data block, correspeedcny book of party bits cast dor ber [0 to party cols © Do Pend uke of a daweed block, cowpute XOR of bis Proce porrespoudeny blocks (iach party block) Brow ober dohe. boat (e) RAID 4: block-interleaved parity ba سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 26:
+ ROW Level (Ovu.) BROW Level & (Coa.) © Proves Kicker VO res Por indepeaded block reads trac Level O ° block read ype too stage disk, sv blocks stored ‏وله من‎ disk cot be reed ict paralet © Provides high trowsPer rotes Por reads oP cauliple blocks frac ‏وی‎ ‎۶ ‏میت ولو‎ 3 black, pony dota cust be ‏لجشحصيت‎ > Cant be doe by veri of pany block, ofl vce oF curred block od xew uche of curred block (© block reads + © bch writes) * Or by reseeepuiteg he party volve ustoy the ce volves oF blocks porrespoendiay te the party blots موی ول ‎oP‏ موی وا مت ۳ موه و() - © Ponty block bevowes u botleuech Por tdepeoded block writes stare every block wore cbse wrties to party dist: سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 27:
+ ROW Level (Ovu.) © ROW Led S: look~‘oierewed Dieiibuied Poriy; portioes dota ced pony coooy ol D+ (disks, roher toc stortoy dota io D disks ood party ta 1 ists. © 05 ‏بشهم بحاصل © كلابب ر.ن.‎ block Por oly set oP bloke te stored va disks (7 wodS) + (, wi the dota blocks stored ‏من‎ the vier P dishes. vit ) RAID 5: block-interleaved distributed parity Deedee Ort Ovwyt- O* Btorn, Ours, OOOO. “or ‏لاس09‎ 

صفحه 28:
+ ROW Level (Ovu.) ‎ROW Leet S (Cou)‏ "ا ‎Aico UO retes tras Level‏ © ‎” @leck ‏ات‎ cazur tt pardlel P the blocks ood their panty blocks are oo ‏او مس‎ ‏را خن اساسا علسه لا را مه ‎Gibsuves bevel ®: provides‏ © ‎tsk.‏ ‎BROW bevel 9: (P+Q Recarkoary svhewe; stator to bevel S, but stores extra recuadda Porat to quand acai wulige cists Pahures. ‎© @eter retdbiiy thea Level Goto higher vost; ot used us widely. ‎ut ‎(g) RAID 6: P + Q redundaney ‎ ‏سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 29:
+ Obove oP ROW Level © Poors ta chovstey ROTO bevel ‏یت و‎ cost © Perhorwaue: Ouwber oP VO opersivas ‏سم له اجه چم‎ ‏وه لو بل‎ ۱ ‏ال لاه چاه تانب رید وا و‎ * ‏لا وا مه سود ملع اص‎ Potted chests: BROW O & eed ody whe data soPey ‏او سح‎ © ‏بیع‎ dota coo be recovered quickly ‏اه مس‎ sources © beet 6 on @ cover wed stu they oe subsuved by 9 ord S Bi bev 0 & oot ‏وا و بو رو وا سوه وروت لس‎ reads ty aoorss di deke, waster deb ana wavered, whick block piri (evel S) works © bed 0 & reel wed ste bevel (od S oPPer adequate sch ey Por okvost ol pplication © Go cowprtion & betwee ) ‏بد © لج‎ سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 30:
+ Oboe oP ROW Level (Ovu.) bevel 0 provides work better write perPorewaare tho level S © bevel 9 requires ot leust © block reads ond © block wries to write a sie beck, whereas bevel | voy requires © black writes © bevel ( prePerred Por high updite eoviroemecis suck as log dhe © Level ( kad higher storage covet hoc level G © ok deve mention ‏سوه‎ reply (GO %lyeu,) wherew dek worse ‏ويم‎ ‏لبط سم‎ wack bse (© 1 ID peors) © VO requireweds howe kerewed wedly, e.g. Por Deb servers © Oke eaugh disks hove bero bough to sutsPy required rate oP WO, they oPtet ] copoly ۱ go here is oPted ow extra worry vet Por bevel (I! B Level S ts preferred Por opplouiows wih buy update rate, red forge ‏سای‎ oF coat B® Level (ts prePerred Por ol wher ‏او‎ سا0 لح 0 لا سواه 1 ههه ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 31:
+ Arrives www © GPwoe ROW: ROTO topewectivas doce coins is scPvore, wit 7 spevid ‏اوه وله‎ ۲ ‏ولا‎ ROW: ۵1۵ ‏وه او لت مه‎ ۱ © @ewore: power Palure dunt wrte coc result corrupted disk ١ ‏بر‎ Pouce Per varies coe block but bePore writes the seozed tao ‏لي‎ ۱ ‏اس‎ corrupted dota ust be deterted wheo power ts restored © Recovery Prox comupiig is sthar to recovery Prox Patled cist: ~ OO-ROO helps to ePRideciy detevied poteatdhy corrupted blacks » Otkerwitse oll blocks oP disk o7ust be ‏لاس وی ای ام‎ ‏رما‎ blocks: Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ea ©Sbervehnts, Cork ced Cnakershe 

صفحه 32:
© Wot swoppieg: replceweot oP disk while sysiew ts rosa, uwihout power dow © Gapped by sue harkwore ROD ‏مود‎ ‏با ول‎ tp revovery, ood keoproves ‏ساره‎ © Oxny systews wotdtat syore disks whick ore kept voter, ood used os replacers Por Rated disks ‏ات خن مس مه الم‎ ارحص صا دا ‎Reduces‏ © مهب نطو( و مر و و فا مه وود 3۹۱۵) رل ‎Oey‏ © ‎the Pucntizaiog of the systec by veto‏ © Reduedeal power supplies wih beter ‏وا‎ ای امن و سپس ‎Ouhile coeivlers oad wutiple:‏ © لصوم 1 Ovweyte- O* Btn, Ours, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 33:
Bl ‏ام مرول‎ oly woecrery (COAROM) © Reward debe, OPO OB per deok © Geek toe bint (OO weer (opted read heud te heater unl skwver) © Asher kieary (OOOO RPO) ceed brver cktrrneber race (9-0 Gls) ‏وج موس‎ ches 18 Oxjtal Odeo Disk (DOO) © DOO-S bok FP GO , an DOO-9 tok 0.5 CO © OOOO ad DOO oe double sided Porwvuts wih vapaciies oF OP GO und 1۶ 09 © Glow seek toe, Por sue reusves us OO-ROO Bo Revd owe versie (OAR exnl DOOR) ore popukr © deta roe ody be writen pare, cee corr be erased. ‏واه و لح سا با لت وی نا و‎ store © Ourunte versivar (CO-RO, DOO-RO, DOO+RO ved DOO-RBO) ‏مره مه‎ 1 Ovweyte- O* Btn, Ours, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 34:
Ouqeaio Popes تاو نا علخ له ول خن جارس ‎Bok lage‏ © Pew CO Por OOT (Orgtd Puce Pape) Porat, ID-PO BB ‏اس‎ ‎OUT (Orgad bicear Pope) Porcat, (OO BB+ wil Obtuce Porat, ood 990 CO wt Oxopex kelool soos Port © DreePer rates Prow Pew to (Os of O@ls © Curediy the cheupest storage wedi © Popes we cheap, but ost oP drives ts very hick © Oery stow ures thee fo copra te woysetic dishs ond optical disks © teoted to sequal ares. © Gowe Ports (Povets) provide Poster seek (De of secouds) of ooet of ‏وی سنا‎ © Osed wat) Por bochup, Por storage oP ‏جه له شوه ای همه‎ wn oP Poe edhe Por trace Pertag fPorwaive Pro vor spetew io coher. BL Dope jukeboxes wed Por very kixe oupany stone” © (rahe (10 byes) to persbve (UO bes) Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «oe ©Sbervehnts, Cork ced Cnakershe 

صفحه 35:
جوووو) وم ما ۲ ۵ ‏ما لحم بخ اد‎ Pred-eoghy storage units cule bloke. Blocks ‏لواصم مول له ممتدوما جعصماه كلاصيا خام جاى جه‎ مها وتو ما او ای سا مت و له مرو بادت() ۲ جد رتسا روا تا ان وی با مد مه )موه لو ال یط ‎weeny blocks os possible to‏ © Per — poriva oP ota wewory wwokible ty store copies oP dist blocks. ۲ ‏موی وق‎ — subsystew respousble Por olovatey buPPer spore to wart wewmry. سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 36:
0 ی + ۲ ‏مر‎ cll vo the bubPer comer whe they ueed a blocks Prow dish. 0 OP tke blocks te ohready to the buPPer, buPPer coccger retucas the address of the block te ovata weeny OP the block ts ont tothe buPPer, the buPPer ‏وم‎ ‎\ @llettes space to the bubPer Por the blots: ۱ ‏سام‎ (hrowien of) sox cher block, Breqired, ‏وا‎ ‎space Por he vew bck. OO. Replaced block vomited bark to disk cal Pitas wodPted staze the sweet revedl tems thot vas ‏مامت‎ to/Petched Pro the dish. (Reads the black Prow the disk to the buPPer, ood retiree the ‏سول‎ of he block ta cota swewory te requester. سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 37:
OvPPer-Replwewed Poles Bl Oost opercty systews rephice he beck bast reveuly wed (LORD srt) Idea behied DRO — use post potera oP block rePereuves uso precicior oP Puture ‏وعصمج اس‎ ۲ ‏سوير لسوج واو ما یس‎ puters (such os sequraid soos), ood a dobar ‏وی و وا موق باس موه مرو‎ suey ‏و هط ام وا‎ اه سوه لصو امه وا وه وی ‎a bed strutewy Por‏ با موی ‎DRO‏ © ‎tot‏ چا اه و ‎the pic oP © retiiccs mond s by‏ هجو ‎wheo‏ ریم ‎Por ack tuple i oP do‏ ‎Por pack tuphe te oP =o‏ ی الم یه ‎P the‏ الم وه مساو من طا کیت رو لو( و ‎by the query opivizer te prePerdble‏ Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. “or ©Sbervehnts, Cork ced Cnakershe 

صفحه 38:
+ OvPPor-Repkowwed Pobsee (Orw.) B® Phoced bok — wewory block thot ts ot dlowed to be writes barks to disk. ۲ Doee-howedu sire — Prees the spare ooowpied by a block oe sora 0 the Proof tuple oP thot blocks kas been processed © Ovstrevedly weed (DRO) sirateqy — syste wast pia he back vurrecty ] processed. (PPier the Pied tuple oP thot block hee beeu processed, he Dock vopeeed, ood & boven the west reveal wed bok, Bb Per exnrner oun Wee ptuteibal ‏رم مت و‎ he probubiay thot a request wil rePereore 3 partirdar rebar © @.x,, he dota deiooay & Pree) wceszed. Wewttir: keep dato ‎te coca weeny bubPer‏ ساسا رم ‎۲ ‏موه مه مت له‎ Porved ouput oP blocks Por the purpose of ‎recovery (soore it Chopter (1?) ‏سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 39:
سم 0 + © Vhe dotbuse fs stored os u ovleviva of Files. (Buck Ae is 0 sequeue of records. (0 revord is a sequeue oF Picks. © Ove wren: ۶ ‏مت‎ record size ip Ped © puck Ale has records oP vor ‏رای جاح‎ © dPPered Pleo ore used Por dPPeredi rebaiivos Oiis cose ip easiest to operat, ull ozosider vortable leach records hater. سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 40:
علدت وال6۳ 4 ای( ۲ © Gere record fetter ‏جات اس ,()-) و سوام‎ he ste oF rack لیر و © * OodRicaicd: ‏ول‎ ant glow records te aross block boverdartes Perryridge Round Hill Mianus Downtown Redwood Perryridge Brighton Downtown © Oetetod oF revord 7 ‏هت‎ ۳ و ور وا © ‏و لس ضحم‎ © do wt wove revo, but Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 41:
Cree lew © Gire the oddress of the First deleted revord to the Pie header. Ose this First record to store the address of the seooed deleted record, ued sv oct ۳ Cow thick of these stored akdesses we ponibys oer hey “poral te ‏مسا‎ of record. Bl Qore space ‏و ما‎ reuse space Por word trbtes of Pree ‏لس‎ ty store potsers. (De punters stored kr i-use records.) Perryridge Mianus Downtown Perryridge Downtown Perryridge Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ara ©Sbervehnts, Cork ced Cnakershe 

صفحه 42:
سوم > أب لبون + تعيب اجنود جا هط و ‎ater‏ و بو( 11 هو را ‎Gernp of antipe record‏ © طايخ سس و ‎Record per that dow vartable beaches‏ © © Reoord yee that dow repeat Pek: (ued kr every kde chats wwodels). سا0 لح 0 لا سواه 1 وم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 43:
Block Header Records fi Entries B Gbted pace header ‏:صصص‎ ‎۶ ‏و او ان ای‎ © earl oP Pree spare ie he block © braze ond stay oP park never B® Records coo be woved arcund wihio 0 poye to heey thew ovciiquous ‏جم كلابب‎ ‏روج‎ spore between theo) eoiry io the header cust be updated. § Cottters should ot potat ‏لو وا رام‎ — tostead they should potato the eoiqy Por the record to header. Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «eo ©Sbervehnts, Cork ced Cnakershe 

صفحه 44:
+ 07 ‏دج تي‎ of (Records x Plow ۲ Wew — oreverd coo be paved caywhere ta the Pile where there is spore © Gequeutd — store records iu sequeuitd order, based va the volue oP the search ‏اس که را‎ record ۳ ‏مه له مخ ام و بای را‎ some oirbute of rack record: the rem ‏امد‎ ta whick block of ‏اه لطس‎ be ‏ام‎ Bl Records of rack rektion way be stored io peparde Pe. ‏وافلج + ذا"‎ ‏تحص ات‎ Ale oryacizatoa records of severd dPPered rettiocer can be stored 1 ‏یی بل‎ He ( سوت وا وا سوه سا من سس لول سه مه( ۴ 1 Ovweyte- O* Btn, Ours, OOOO. ‏سا0 لح 0 لا سواه 1 جوم‎ 

صفحه 45:
+ Orqucad Pie Orqatzatva © Guituble Por opphouices thot require sequeutd provessiay of the euire ‏عا‎ LL? ‏مر‎ ‎۳9 ‎LL? ‎۳ ‏سم‎ ‎۳ ‏حم‎ ۲ ۱ records in the Ale oe ordered by ‏راو‎ Brighton Downtown Downtown Mianus Perryridge Perryridge Perryridge Redwood 00 Round Hill Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 46:
تست — ‎Deletion‏ نا ۳ ‏ای ما وا لس سا مان و با مس مه‎ © B here & Pree opae kee here © Pow Pree spare, tsar ‏و او سا‎ over Pw books © ‏اس اه و ی له و‎ updated ۸2۱7 | Brighton [7507 3 Bl Weed ie reorcprtze the le Boas ieee io erase reek A-101 | Downtown | 500 + sequecid order A-110 Downtown 600 = A-215 | Mianus 700 ۳ A-102 | Perryridge [400| ‏د‎ ‎A201 | Perryridge [900] 4 A218 | Peryridge [700] + A222 | Redwood [700] 4 A-305 | Round Hill 350 A-888 | North Town [800 [4 Ez Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «eo ©Sbervehnts, Cork ced Cnakershe JNVVNTINAN 

صفحه 47:
+ Onde Ohsteriog Pie Orqactzuion Gtore severd retizgs to coe Ae usiog 0 aulttoble chustertagy Pie ‏مس‎ customer_name | account_number customer_name | customer_street | customer_city Hayes Main Brooklyn Turner Putnam Stamford Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. wer ©Sbervehnts, Cork ced Cnakershe 

صفحه 48:
+ Ontidble Olhwiertay Pile Orqacizaiva (ovci.) Outitable chustertog orysaizaica oP ‏لمه -وصخاصت‎ depositor! Hayes Hayes Hayes | A-503 © xpd Por quetes ube depostor تبظكاسم«,‎ ued Por queries: Kauch ry oe ‏يواد‎ aptow@er oad hte aso © bn Por ‏و‎ © Cu add potter choke tp hak records oP a parade rebate سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 49:
عووسوت) برومممطظ) مب ۹ ‎suck or‏ رل ات ند 1 ۲ اج سب ۶ ماس ‎of pack‏ سا تاه بلج بت ۶ ‎oF views‏ سا امه سوم ل مس رو 8 ‎Porratiog, forkdkry passwords‏ وی له ها ۲ ی ما ام و ‎of‏ حاسمت © تاه ما ۸ لوا تا ‎Whew rekton & stored (sequectilhash/...)‏ © ما ‎hyped bodied of‏ © ‎Bl ePoreaion chou knices (Okorer (2)‏ Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «.e0 ©Sbervehnts, Cork ced Cnakershe 

صفحه 50:
(عد2) عبعسدا8) ۱ © Cotby structure © Rettiowd represectation ra dob © speondred deta structures deskred Por ePeiect anmese, ki ‏لمحم‎ مس ری ‎posstte‏ 0 ۲ (مصي الجسدمم_لصصررصد بصو _ وص ) = مومس وو( ‎krdex Pe,‏ جمر_ مام صم أصام) ‏ - امود جاه 000 [(صضحاه ا ‎dePiesioa)‏ ب_سض) يبلن همه 4 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 51:
Gad oP Okaper 

صفحه 52:
(Reourd Represecicipa BE Rewords wi Pico bears Pek ‏سوم صا برت جه‎ ۶ ‏لوق‎ tp records (sini) fo progres KRRRRRE ۶ ‏له مس سس‎ ches ١ Bax. a bisa kerk ‏ای اب‎ are ‏اه‎ ۳ ‏ما لبون‎ Pekbs con be represeuted by «pct (oP Poet lech) kere oP Peet te the locotica uihic the record ord feces ‏لماعتا جا‎ feces. © Ol Betts stot of predePiced loruicg, but extra tedrevioa required Por سوت vortable leony Pektes 7 Law | F00 | سا 1 ميهي اميا 009 امه مه و «0.00 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 53:
Perryridge Round Hill Mianus Downtown Redwood Perryridge Brighton Downtown Perryridge «0.00 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 54:
۸ Cle oP Pique 00.9, wik Revord C Deleted aad Ol QRevords Doved Perryridge Round Hill Downtown Redwood Perryridge Brighton Downtown Perryridge سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 55:
fe Cle oP Pique 00.9, Otk Revord © dobtied wad Pid + QRevord Ooved Perryridge Round Hill Perryridge Downtown Redwood Perryridge Brighton Downtown Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «0.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 56:
+ ®yte-Oriaq Repreveddiva oP Owtdble-Leagk Qevords ۸-8 900 600 «0.00 400 350, 700 500 700 750 ۸-2 ۸5 ۸5 ۸-1 ۸22 ۸7 ۳ Round Hill Mianus Downtown Redwood Brighton ات شر دم ين ص ين 000 Overy - O* Ot, Our 2, 

صفحه 57:
1 Ovweyte- O* Btn, Ours, OOOO. wor ©Sbervehnts, Cork ced Cnakershe 

صفحه 58:
Brooklyn Putnam | Stamford A-305 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. ‏موم‎ 

صفحه 59:
account-number A-102 A-220 A-503 A-305 customer-name Hayes Hayes Hayes Turner Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 68 

صفحه 60:
customer-city Brooklyn Stamford customer-street Main Putnam موه customer-name Hayes Turner Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 61:
Stamford A-503 Putnam Hayes Turner | Turner | A-305 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 62:
(@) RAID 3: bit interleaved parity (¢) RAID 4 blockinterleaved parity wists (RAND 5: block-interleaved. ۱ (g) RAID 6 P + Q redundancy 

صفحه 63:
0 9 1 Btorn, Our, OOOO. 

صفحه 64:
+ ‏مج‎ 00. A-102 | Perryridge | 400 A-305 | Round Hill | 35 A-101 Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. woe ©Sbervehnts, Cork ced Cnakershe 

صفحه 65:
۸۵-102 | Perryridge | 40 Round Hill Perryridge 0 500 700 A201 | Perryridge | 900] [aio | Downtown | 600 سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 66:
48) RAID 0 +1 with a single dik face سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 67:
+ ۳ 0 Brooklyn Stamford «aor ۸-03 Putnam ۸-5 1 Ovweyte- O* Btn, Ours, OOOO. 

صفحه 68:
Perryridge | A-102 Round Hill | 5 Mianus A215 Downtown | A-101 Redwood [| A-222 Brighton _[ A-217 yee siren represen rack «ns exntet record (L) ‏لحب مما بد همان هچ‎ oP each record ‏وان‎ uth delet (PPro wks rons سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 69:
+ ‏لاس۳‎ Reprrorctaica © Ose ove or wre ‏تاحصم را ات‎ © reserved spare © potters § Qeserved spore — cao use Pixeddeayh records of ‏موی سا و‎ leagh; uused space te shorter nevords Pled wah a oul or ‏لمحتو لمج‎ Perryridge Round Hill Mianus Downtown Redwood Brighton سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 70:
Perryridge Round Hill Mianus Downtown Redwood Brighton © Poitier wetwd © 0 vorttbleteuyh record is represroted by a lst oP Pred-eagh records, و لها لت © Con be wed veo 8 ‏مومت سل‎ revord leagh is wt ker Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 71:
+ Porter Dobe (Orc) § Oisadvoctaye te pricier structure; spare is wasted ta dl records اه و و ام ام B Gokiica is to low tie brads of blocks tet Pte: © ‏یط‎ block — pootoies the Pirst records oP choir © QuerPlow block — eoetaes revords vier thoa those thot are the 400 + 350 700 500. 700 750 + 900 م 700 ۸-2 8-305 A-215 ۸-1 ۸222 A-217 A-201 A218 ۸0 ae First records oP chairs. fnchor | Perryridge Hock Round Hill Mianus Downtown Redwood Brighton loverflow block Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 72:
+ ‏بان ذاه وداج(‎ to Piles ۲ ‏ماه پم(‎ to Ales ip skthor ty woppien tuples to Ales ‏موه لا و و‎ object date coc be stored usta Pie structures. 11 Obert in O-O ‏وال‎ way hack uaPoreiy oad aay be very barge; such jens ‏سوه اما و و و تخل اج وا سم‎ © ‏ری لام سا رو ماه اه چاه مره تب ۳ بو‎ data structures suck us ‏عطا لها‎ © Get Rebs with o harger ‏ماو ها روت وه اه این‎ os separa velticas te the database. © Get Pelle coc dbo be elcvicaied ‏اه بو اس موه سا و‎ ١ Greolar to ooaversica oF ulivdued utidbutes oP (B-R disgrace to rebates Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «0.76 ©Sbervehnts, Cork ced Cnakershe 

صفحه 73:
+ Drppray oP Obpue to Ply (Ove) 9 Obreoe we kkeciPed by wt oben NeuiPier (OD); he sirroge syste ured ‏و‎ ‏اه مه سا وا مه‎ ed its OD (fis anton tr rled derePereactoy). © ‏حيط‎ KeuiPers de ont dreviy spechy oc oben!’ physic location; oust ‏وه موی‎ todes thot wraps ot OD ty the vbjert's ‏ما امه‎ (ede hr eto he eb the ‏سم مام‎ be Bar et ‏زاس‎ Physic Os typirdly have the ‏مسا‎ ‏مس‎ or Fle tector ©. apne HleoiRer waka fhe ‏ماس‎ or Pe 9. ‏من‎ oP Peet wikia he poe Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «0.76 ©Sbervehnts, Cork ced Cnakershe 

صفحه 74:
+ Dannpusd ob (Persia Pokies 15 ‏ه سا نوی 0102 تسوا‎ vcique KeulPer. This their stored ts the shart cee cent essed an ‏مسجو سمج جر مجوو‎ borden peer. Physical Object Identifier ‏تماد م‎ 519.56850.1200 ] 51 Volume Block Offset | Unique-Id Good OID [519.56850.1200 Data Bad OID [19,56850.1200 (a) General structure, b) Example of use, سا0 لح 0 لا سواه 1 و ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 75:
+ Qacnewed oP Persstiea Prides (Ovd.) ۲ hopewed persisted priders ust O10s; persistent potters are substratchy ‏و روموت و ما متا‎ Potter sutzzhey ‏له‎ dowe oo cost oF lovatey persisted! objects dread ie wewory. © GoPwere swizztey (swizzhoy oo potdier dePereue) © Ohea a persisted porter t& Prot derePerewed, tee ponte ‏ییاه‎ ‎(repkoed by ou keener) porter) Wer he oqo! & bouesl es wer. © Gubsequeci derePereues of oP the sexve poicter beoowe cheap. © Phe physic located of oa oben! tr wewory wet ot choage P swirled Potters pout te it) the srlution ts to pio poses ‏تچ و‎ © Okes oo oben ts vorites back ty disk, coy suizaled poitiers ۱ ‏اه وه‎ to be ‏ری‎ سا0 لح 0 لا سواه 1 هوه ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 76:
AWardwore Guszalay © ik herders swizzley, persisted potters ta objevts eed the sue xneant oF spare os Kee wory potiers — extra storage ‏و اجه‎ the ‏ها متام‎ weed to store rest of potter ‏تا‎ © Oses vitud wewory trowstaiva werhovisw ‏الم اه وا‎ ‏متا ویو‎ persisted potuers und ‏.ججكهامم تم‎ ۲ Ol persisted poitiers ‏وه و و و‎ swizaled whe the pays if Pst read ic. © tes proyrniwers hove te work wil jvet coe type oF potter, te, ie expo) potter. © secre oP the swirled potions ‏و ری ام وا رو‎ hot ore curred oot domed oxy ‏موه امم جل او رم لو‎ und deta) Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «0.26 ©Sbervehnts, Cork ced Cnakershe 

صفحه 77:
+ Areinee Quick Bersted porter ib oroepinly pl io tw ports! a pace tealPer, ond or oP Poet wakia the poe. © De pace tlewiPer too potcter ts 0 short tedrent potter: (Book poxe kas 3 ‏متسد‎ table hel provides: 9 svuppiey Brow the shod pays Meiers to Pull ‏ون لول‎ KleuiPiers. © tobe Por 3 pace ts scr (ot wrost IDCP porters tao POSO bye pace uk @ byte pokior) © Outil potters ta pace i the scx pare shore sexe ‏رو‎ te the ‏ما‎ ‎ible Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. are ©Sbervehnts, Cork ced Cnakershe 

صفحه 78:
(0) واسه6 شم 170 object 3 Pageld Off. 020 2395 Pageld Off. Pageld Off. 2395 ] 5 4867 object 1 object 2 PageID FullPageID 2395 | 679.34278 4867 |] 0 translation table Bl Pane keane bePore swtzatog (pace bodted ou ctoh) Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «0.20 

صفحه 79:
+ ‏سس © وطسرا"‎ (Orn) لاه ‎tn the poye ure‏ وم موم باتوی وا ون و لوا مور جدا(0 © ‎desorbed below‏ جد 6 Persietedt potters i eark ober ia he paye are booed wei oben! ype KPoremaion ‏ات و‎ persisted portion (p11) Piatt Pll pare 1D (P, ۱ APO, ches voldrenh hwe ‏و مه را نا اسر سوم ری لعج و‎ urd cmxvory pase (Pond ‏موی بط سول‎ » Wore! here ceed wt be cay physiod spare (wheker ‏موی وا‎ or oa cok surp-space) dlocated Por ke ‏حا ب جومم مه لعج‎ port. Space oat be ‏سا لس‎ F (oad when) (Pie aces. “la his cose ‏سول‎ ‎wt required. Oovessie, a wewory lovalivs io he page is the wil result io seqerectaion Uioktiog, udbick te hooded oe described hater Let v, be te uri pace corded iP, (ether ewer or cove) ‏تم لت رمع‎ ‏را روت ات یه‎ P,) ia he roehtion bk, by (v, P,) Ez Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. «0.20 ©Sbervehnts, Cork ced Cnakershe 

صفحه 80:
(0) واسه6 شم ‎detects‏ و رو سا رس و روم مه سانا و سوام و موه لح ما ‎the pene ponte to hes oot‏ ‎DorUrs.‏ مات مود ‎oper) o Ructon ty be kwoked oo‏ جا لج ‎is Dots‏ ۱ یا ‎severe‏ ‎Bl Phe Rowton doer the Polk when tt Reohed‏ ‎(sua-spare) Ror he page cocks he rePerecred‏ موه تما ‎hos ot ber domed rater. Pura off reat‏ موه سای ار ‎(Read ta the pace Proc cho‏ ‎(Perfor pointer suvtzziey Por cack persisted potdter ic the poe, 0 desorbed carter ‏سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 81:
Pageld Off.| | Pageld Off. 4867 1020| [so01 [170 object 2 object 3 PagelD FullPagelD 5001 | 679.34278 4867 | 519.56850 Pageld Off, 5001 255 object 1 translation table age koa Per ‏شوه‎ © Page wih short pre teuiPier COOS wer ulovcted address GOO. Observe chooge to potters oad trocstatios table. © Page wih short page deter POO? hes beeo dlocuted address ۵9۶, Or chase ta potter oad tremetatica tobe. Deedee Orie ‏مسف “9 - عابس‎ Ours, OOOO. 

صفحه 82:
+ ‏سس © وطسرا"‎ (Orn) اوه له موی مش وا ام وال ون باه له رتیه © جععدم رل متسه ۳ و( ‎aware thot thas persisted poiters,‏ مه به و ماه سا ره مس( و ‎coy way!‏ و له ع نا له ‎wand do oot‏ ‏و رس با و له ‎Teer exten ode‏ مه سوه مه مه او بط رب بط تا ۳ سین ۱ ‎Kieu, 0‏ 0 |[ | ز[ز[ز | ز ز ز [ز ز ز 0 0 مرک ۳ و اإسجم سا نا تقو مات ‎mort be saved certs ty choke‏ جوم الط و ‎We ceed‏ © ‎© @ set of ‏متام بسن سوه موه موی موم‎ bbe, Pages ox oil be swizakal ay wend whew Perched (ok ovpy oP rxnehatoa tubke ceeded). = Om fond cot access wore pes thon size of ‏سا رو لعج‎ of vind weeny adresses Por rher paws t exper ‏سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ ‎ 

صفحه 83:
© ‏وچ و اوه و اه ای و مب با‎ way be dPPered Prow the ‏ات و لو(‎ hey ore stored go disk to ‏وه )لول‎ © soPivore swizzhay — siruture oP perssted od kewewory poniers are ‏سل‎ ۱ dPPered dat ‏جكساد سور‎ اه موی ال اج ‎Dche the physicd represrutaiva oP objevts‏ )© ای با له لو © Cag ‏صصص‎ cower Brow dek represectaton to Pore required oo hor spent wurkwe, henna, ox reopen, whe te obec (or ‏تحص با(‎ kalo wecoory. سا0 لح 0 لا سواه 1 66 ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 84:
+ bore Obprs BE bone obits : bray hare obfeote (bobs) onl character karop objects (cbs) © Grenoples keke: ۱ ‏سم مب‎ ۱ ‏ری‎ data suck os Keages ‏و له موی له‎ cud ood vider cats bore bev way ceed ty be stored ‏هه صصخت و و‎ oP bytes wheat brought inte wewory. © 1P ca vbien is bayer theo o poe, ouikquous payer oF the bubPer pool west ‏.ةا یه وا واه با‎ © Dey be prePeruble to dodo dent uovess to dota, ood oly low ores through 0 Ple-systeutike (OP 1, to rewove oeed Por ovuicuaus stone. سا0 لح 0 لا سواه 1 موم ‎Ours, OOOO.‏ مسف “9 - عابس ‎Deedee Orie‏ 

صفحه 85:
+ DodP yim Lar Oboe سره وه ‎requires teseriidelete oP bytes Pow speviied reqvus oP‏ ل لا © Otter Pie oryacizaiva (desorbed hater ta Chapter (@) coc be woodPied to ‏اه ها ام‎ © ‏سا اس‎ pore of the tree stores betwee oP ced (| pace work oP dota Pow the ober © Gpecdtpurpese ‏ماو‎ proqraws puiside the dotobase ore ‏جلج وت و لح‎ axe obec: © Der dota rected os 0 bye sires woaiputied by ediors oad Porcvaters. © Grophicd dota ood oude/vider dota is ppicdly credied oad deployed by separaie ‏ددص ابوج‎ عون اه مه اجه او بوحصجوی و( ‎checkoulcheckta wetbod‏ © 1 Ovweyte- O* Btn, Ours, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

Chapter 11: Storage and File Structure Database System Concepts, 5th Ed. ©Silberschatz, Korth and Sudarshan See www.db-book.com for conditions on re-use Chapter 11: Storage and File Structure  Overview of Physical Storage Media  Magnetic Disks  RAID  Tertiary Storage  Storage Access  File Organization  Organization of Records in Files  Data-Dictionary Storage  Storage Structures for Object-Oriented Databases Database System Concepts - 5th Edition, Aug 12, 2005. 11.2 ©Silberschatz, Korth and Sudarshan Classification of Physical Storage Media  Speed with which data can be accessed  Cost per unit of data  Reliability   data loss on power failure or system crash  physical failure of the storage device Can differentiate storage into:  volatile storage: loses contents when power is switched off  non-volatile storage:   Contents persist even when power is switched off. Includes secondary and tertiary storage, as well as batter-backed up main-memory. Database System Concepts - 5th Edition, Aug 12, 2005. 11.3 ©Silberschatz, Korth and Sudarshan Physical Storage Media  Cache – fastest and most costly form of storage; volatile; managed by the computer system hardware.  Main memory:  fast access (10s to 100s of nanoseconds; 1 nanosecond = 10 –9 seconds)  generally too small (or too expensive) to store the entire database   capacities of up to a few Gigabytes widely used currently  Capacities have gone up and per-byte costs have decreased steadily and rapidly (roughly factor of 2 every 2 to 3 years) Volatile — contents of main memory are usually lost if a power failure or system crash occurs. Database System Concepts - 5th Edition, Aug 12, 2005. 11.4 ©Silberschatz, Korth and Sudarshan Physical Storage Media (Cont.)  Flash memory  Data survives power failure  Data can be written at a location only once, but location can be erased and written to again  Can support only a limited number (10K – 1M) of write/erase cycles.  Erasing of memory has to be done to an entire bank of memory  Reads are roughly as fast as main memory  But writes are slow (few microseconds), erase is slower  Cost per unit of storage roughly similar to main memory  Widely used in embedded devices such as digital cameras  Is a type of EEPROM (Electrically Erasable Programmable ReadOnly Memory) Database System Concepts - 5th Edition, Aug 12, 2005. 11.5 ©Silberschatz, Korth and Sudarshan Physical Storage Media (Cont.)  Magnetic-disk  Data is stored on spinning disk, and read/written magnetically  Primary medium for the long-term storage of data; typically stores entire database.  Data must be moved from disk to main memory for access, and written back for storage  Much slower access than main memory (more on this later)  direct-access – possible to read data on disk in any order, unlike magnetic tape  Hard disks vs floppy disks  Capacities range up to roughly 400 GB currently   Much larger capacity and cost/byte than main memory/flash memory  Growing constantly and rapidly with technology improvements (factor of 2 to 3 every 2 years) Survives power failures and system crashes  disk failure can destroy data, but is rare Database System Concepts - 5th Edition, Aug 12, 2005. 11.6 ©Silberschatz, Korth and Sudarshan Physical Storage Media (Cont.)  Optical storage  non-volatile, data is read optically from a spinning disk using a laser  CD-ROM (640 MB) and DVD (4.7 to 17 GB) most popular forms  Write-one, read-many (WORM) optical disks used for archival storage (CDR, DVD-R, DVD+R)  Multiple write versions also available (CD-RW, DVD-RW, DVD+RW, and DVD-RAM)  Reads and writes are slower than with magnetic disk  Juke-box systems, with large numbers of removable disks, a few drives, and a mechanism for automatic loading/unloading of disks available for storing large volumes of data Database System Concepts - 5th Edition, Aug 12, 2005. 11.7 ©Silberschatz, Korth and Sudarshan Physical Storage Media (Cont.)  Tape storage  non-volatile, used primarily for backup (to recover from disk failure), and for archival data  sequential-access – much slower than disk  very high capacity (40 to 300 GB tapes available)  tape can be removed from drive  storage costs much cheaper than disk, but drives are expensive  Tape jukeboxes available for storing massive amounts of data  hundreds of terabytes (1 terabyte = 109 bytes) to even a petabyte (1 petabyte = 1012 bytes) Database System Concepts - 5th Edition, Aug 12, 2005. 11.8 ©Silberschatz, Korth and Sudarshan Storage Hierarchy Database System Concepts - 5th Edition, Aug 12, 2005. 11.9 ©Silberschatz, Korth and Sudarshan Storage Hierarchy (Cont.)  primary storage: Fastest media but volatile (cache, main memory).  secondary storage: next level in hierarchy, non-volatile, moderately fast access time   also called on-line storage  E.g. flash memory, magnetic disks tertiary storage: lowest level in hierarchy, non-volatile, slow access time  also called off-line storage  E.g. magnetic tape, optical storage Database System Concepts - 5th Edition, Aug 12, 2005. 11.10 ©Silberschatz, Korth and Sudarshan Magnetic Hard Disk Mechanism NOTE: Diagram is schematic, and simplifies the structure of actual disk drives Database System Concepts - 5th Edition, Aug 12, 2005. 11.11 ©Silberschatz, Korth and Sudarshan Magnetic Disks   Read-write head  Positioned very close to the platter surface (almost touching it)  Reads or writes magnetically encoded information. Surface of platter divided into circular tracks      Over 50K-100K tracks per platter on typical hard disks Each track is divided into sectors.  A sector is the smallest unit of data that can be read or written.  Sector size typically 512 bytes  Typical sectors per track: 500 (on inner tracks) to 1000 (on outer tracks) To read/write a sector  disk arm swings to position head on right track  platter spins continually; data is read/written as sector passes under head Head-disk assemblies  multiple disk platters on a single spindle (1 to 5 usually)  one head per platter, mounted on a common arm. Cylinder i consists of ith track of all the platters Database System Concepts - 5th Edition, Aug 12, 2005. 11.12 ©Silberschatz, Korth and Sudarshan Magnetic Disks (Cont.)   Earlier generation disks were susceptible to head-crashes  Surface of earlier generation disks had metal-oxide coatings which would disintegrate on head crash and damage all data on disk  Current generation disks are less susceptible to such disastrous failures, although individual sectors may get corrupted Disk controller – interfaces between the computer system and the disk drive hardware.  accepts high-level commands to read or write a sector  initiates actions such as moving the disk arm to the right track and actually reading or writing the data  Computes and attaches checksums to each sector to verify that data is read back correctly  If data is corrupted, with very high probability stored checksum won’t match recomputed checksum  Ensures successful writing by reading back sector after writing it  Performs remapping of bad sectors Database System Concepts - 5th Edition, Aug 12, 2005. 11.13 ©Silberschatz, Korth and Sudarshan Disk Subsystem  Multiple disks connected to a computer system through a controller   Controllers functionality (checksum, bad sector remapping) often carried out by individual disks; reduces load on controller Disk interface standards families  ATA (AT adaptor) range of standards  SATA (Serial ATA)  SCSI (Small Computer System Interconnect) range of standards  Several variants of each standard (different speeds and capabilities) Database System Concepts - 5th Edition, Aug 12, 2005. 11.14 ©Silberschatz, Korth and Sudarshan Performance Measures of Disks  Access time – the time it takes from when a read or write request is issued to when data transfer begins. Consists of:  Seek time – time it takes to reposition the arm over the correct track.  Average seek time is 1/2 the worst case seek time. – Would be 1/3 if all tracks had the same number of sectors, and we ignore the time to start and stop arm movement   Rotational latency – time it takes for the sector to be accessed to appear under the head.    4 to 10 milliseconds on typical disks Average latency is 1/2 of the worst case latency. 4 to 11 milliseconds on typical disks (5400 to 15000 r.p.m.) Data-transfer rate – the rate at which data can be retrieved from or stored to the disk.  25 to 100 MB per second max rate, lower for inner tracks  Multiple disks may share a controller, so rate that controller can handle is also important  E.g. ATA-5: 66 MB/sec, SATA: 150 MB/sec, Ultra 320 SCSI: 320 MB/s  Fiber Channel (FC2Gb): 256 MB/s Database System Concepts - 5th Edition, Aug 12, 2005. 11.15 ©Silberschatz, Korth and Sudarshan Performance Measures (Cont.)  Mean time to failure (MTTF) – the average time the disk is expected to run continuously without any failure.  Typically 3 to 5 years  Probability of failure of new disks is quite low, corresponding to a “theoretical MTTF” of 500,000 to 1,200,000 hours for a new disk   E.g., an MTTF of 1,200,000 hours for a new disk means that given 1000 relatively new disks, on an average one will fail every 1200 hours MTTF decreases as disk ages Database System Concepts - 5th Edition, Aug 12, 2005. 11.16 ©Silberschatz, Korth and Sudarshan Optimization of Disk-Block Access   Block – a contiguous sequence of sectors from a single track  data is transferred between disk and main memory in blocks  sizes range from 512 bytes to several kilobytes  Smaller blocks: more transfers from disk  Larger blocks: more space wasted due to partially filled blocks  Typical block sizes today range from 4 to 16 kilobytes Disk-arm-scheduling algorithms order pending accesses to tracks so that disk arm movement is minimized  elevator algorithm : move disk arm in one direction (from outer to inner tracks or vice versa), processing next request in that direction, till no more requests in that direction, then reverse direction and repeat Database System Concepts - 5th Edition, Aug 12, 2005. 11.17 ©Silberschatz, Korth and Sudarshan Optimization of Disk Block Access (Cont.)  File organization – optimize block access time by organizing the blocks to correspond to how data will be accessed  E.g. Store related information on the same or nearby cylinders.  Files may get fragmented over time   E.g. if data is inserted to/deleted from the file  Or free blocks on disk are scattered, and newly created file has its blocks scattered over the disk  Sequential access to a fragmented file results in increased disk arm movement Some systems have utilities to defragment the file system, in order to speed up file access Database System Concepts - 5th Edition, Aug 12, 2005. 11.18 ©Silberschatz, Korth and Sudarshan Optimization of Disk Block Access (Cont.)  Nonvolatile write buffers speed up disk writes by writing blocks to a non-volatile RAM buffer immediately  Non-volatile RAM: battery backed up RAM or flash memory    Controller then writes to disk whenever the disk has no other requests or request has been pending for some time  Database operations that require data to be safely stored before continuing can continue without waiting for data to be written to disk  Writes can be reordered to minimize disk arm movement Log disk – a disk devoted to writing a sequential log of block updates   Even if power fails, the data is safe and will be written to disk when power returns Used exactly like nonvolatile RAM  Write to log disk is very fast since no seeks are required  No need for special hardware (NV-RAM) File systems typically reorder writes to disk to improve performance  Journaling file systems write data in safe order to NV-RAM or log disk  Reordering without journaling: risk of corruption of file system data Database System Concepts - 5th Edition, Aug 12, 2005. 11.19 ©Silberschatz, Korth and Sudarshan RAID  RAID: Redundant Arrays of Independent Disks    disk organization techniques that manage a large numbers of disks, providing a view of a single disk of  high capacity and high speed by using multiple disks in parallel, and  high reliability by storing data redundantly, so that data can be recovered even if a disk fails The chance that some disk out of a set of N disks will fail is much higher than the chance that a specific single disk will fail.  E.g., a system with 100 disks, each with MTTF of 100,000 hours (approx. 11 years), will have a system MTTF of 1000 hours (approx. 41 days)  Techniques for using redundancy to avoid data loss are critical with large numbers of disks Originally a cost-effective alternative to large, expensive disks  I in RAID originally stood for ``inexpensive’’  Today RAIDs are used for their higher reliability and bandwidth.  The “I” is interpreted as independent Database System Concepts - 5th Edition, Aug 12, 2005. 11.20 ©Silberschatz, Korth and Sudarshan Improvement of Reliability via Redundancy  Redundancy – store extra information that can be used to rebuild information lost in a disk failure  E.g., Mirroring (or shadowing)  Duplicate every disk. Logical disk consists of two physical disks.  Every write is carried out on both disks   Reads can take place from either disk If one disk in a pair fails, data still available in the other  Data loss would occur only if a disk fails, and its mirror disk also fails before the system is repaired – Probability of combined event is very small »  Except for dependent failure modes such as fire or building collapse or electrical power surges Mean time to data loss depends on mean time to failure, and mean time to repair  E.g. MTTF of 100,000 hours, mean time to repair of 10 hours gives mean time to data loss of 500*106 hours (or 57,000 years) for a mirrored pair of disks (ignoring dependent failure modes) Database System Concepts - 5th Edition, Aug 12, 2005. 11.21 ©Silberschatz, Korth and Sudarshan Improvement in Performance via Parallelism  Two main goals of parallelism in a disk system: 1. Load balance multiple small accesses to increase throughput 2. Parallelize large accesses to reduce response time.  Improve transfer rate by striping data across multiple disks.  Bit-level striping – split the bits of each byte across multiple disks  In an array of eight disks, write bit i of each byte to disk i.  Each access can read data at eight times the rate of a single disk.  But seek/access time worse than for a single disk   Bit level striping is not used much any more Block-level striping – with n disks, block i of a file goes to disk (i mod n) + 1  Requests for different blocks can run in parallel if the blocks reside on different disks  A request for a long sequence of blocks can utilize all disks in parallel Database System Concepts - 5th Edition, Aug 12, 2005. 11.22 ©Silberschatz, Korth and Sudarshan RAID Levels  Schemes to provide redundancy at lower cost by using disk striping combined with parity bits  Different RAID organizations, or RAID levels, have differing cost, performance and reliability characteristics  RAID Level 0: Block striping; non-redundant.  Used in high-performance applications where data lost is not critical.  RAID Level 1: Mirrored disks with block striping  Offers best write performance.  Popular for applications such as storing log files in a database system. Database System Concepts - 5th Edition, Aug 12, 2005. 11.23 ©Silberschatz, Korth and Sudarshan RAID Levels (Cont.)  RAID Level 2: Memory-Style Error-Correcting-Codes (ECC) with bit striping.  RAID Level 3: Bit-Interleaved Parity  a single parity bit is enough for error correction, not just detection, since we know which disk has failed  When writing data, corresponding parity bits must also be computed and written to a parity bit disk  To recover data in a damaged disk, compute XOR of bits from other disks (including parity bit disk) Database System Concepts - 5th Edition, Aug 12, 2005. 11.24 ©Silberschatz, Korth and Sudarshan RAID Levels (Cont.)   RAID Level 3 (Cont.)  Faster data transfer than with a single disk, but fewer I/Os per second since every disk has to participate in every I/O.  Subsumes Level 2 (provides all its benefits, at lower cost). RAID Level 4: Block-Interleaved Parity; uses block-level striping, and keeps a parity block on a separate disk for corresponding blocks from N other disks.  When writing data block, corresponding block of parity bits must also be computed and written to parity disk  To find value of a damaged block, compute XOR of bits from corresponding blocks (including parity block) from other disks. Database System Concepts - 5th Edition, Aug 12, 2005. 11.25 ©Silberschatz, Korth and Sudarshan RAID Levels (Cont.)  RAID Level 4 (Cont.)  Provides higher I/O rates for independent block reads than Level 3  block read goes to a single disk, so blocks stored on different disks can be read in parallel  Provides high transfer rates for reads of multiple blocks than no-striping  Before writing a block, parity data must be computed   Can be done by using old parity block, old value of current block and new value of current block (2 block reads + 2 block writes)  Or by recomputing the parity value using the new values of blocks corresponding to the parity block – More efficient for writing large amounts of data sequentially Parity block becomes a bottleneck for independent block writes since every block write also writes to parity disk Database System Concepts - 5th Edition, Aug 12, 2005. 11.26 ©Silberschatz, Korth and Sudarshan RAID Levels (Cont.)  RAID Level 5: Block-Interleaved Distributed Parity; partitions data and parity among all N + 1 disks, rather than storing data in N disks and parity in 1 disk.  E.g., with 5 disks, parity block for nth set of blocks is stored on disk (n mod 5) + 1, with the data blocks stored on the other 4 disks. Database System Concepts - 5th Edition, Aug 12, 2005. 11.27 ©Silberschatz, Korth and Sudarshan RAID Levels (Cont.)  RAID Level 5 (Cont.)  Higher I/O rates than Level 4.    Block writes occur in parallel if the blocks and their parity blocks are on different disks. Subsumes Level 4: provides same benefits, but avoids bottleneck of parity disk. RAID Level 6: P+Q Redundancy scheme; similar to Level 5, but stores extra redundant information to guard against multiple disk failures.  Better reliability than Level 5 at a higher cost; not used as widely. Database System Concepts - 5th Edition, Aug 12, 2005. 11.28 ©Silberschatz, Korth and Sudarshan Choice of RAID Level  Factors in choosing RAID level  Monetary cost  Performance: Number of I/O operations per second, and bandwidth during normal operation  Performance during failure  Performance during rebuild of failed disk   Including time taken to rebuild failed disk RAID 0 is used only when data safety is not important  E.g. data can be recovered quickly from other sources  Level 2 and 4 never used since they are subsumed by 3 and 5  Level 3 is not used anymore since bit-striping forces single block reads to access all disks, wasting disk arm movement, which block striping (level 5) avoids  Level 6 is rarely used since levels 1 and 5 offer adequate safety for almost all applications  So competition is between 1 and 5 only Database System Concepts - 5th Edition, Aug 12, 2005. 11.29 ©Silberschatz, Korth and Sudarshan Choice of RAID Level (Cont.)   Level 1 provides much better write performance than level 5  Level 5 requires at least 2 block reads and 2 block writes to write a single block, whereas Level 1 only requires 2 block writes  Level 1 preferred for high update environments such as log disks Level 1 had higher storage cost than level 5  disk drive capacities increasing rapidly (50%/year) whereas disk access times have decreased much less (x 3 in 10 years)  I/O requirements have increased greatly, e.g. for Web servers  When enough disks have been bought to satisfy required rate of I/O, they often have spare storage capacity    so there is often no extra monetary cost for Level 1! Level 5 is preferred for applications with low update rate, and large amounts of data Level 1 is preferred for all other applications Database System Concepts - 5th Edition, Aug 12, 2005. 11.30 ©Silberschatz, Korth and Sudarshan Hardware Issues  Software RAID: RAID implementations done entirely in software, with no special hardware support  Hardware RAID: RAID implementations with special hardware  Use non-volatile RAM to record writes that are being executed  Beware: power failure during write can result in corrupted disk  E.g. failure after writing one block but before writing the second in a mirrored system  Such corrupted data must be detected when power is restored – Recovery from corruption is similar to recovery from failed disk – NV-RAM helps to efficiently detected potentially corrupted blocks » Otherwise all blocks of disk must be read and compared with mirror/parity block Database System Concepts - 5th Edition, Aug 12, 2005. 11.31 ©Silberschatz, Korth and Sudarshan Hardware Issues (Cont.)   Hot swapping: replacement of disk while system is running, without power down  Supported by some hardware RAID systems,  reduces time to recovery, and improves availability greatly Many systems maintain spare disks which are kept online, and used as replacements for failed disks immediately on detection of failure   Reduces time to recovery greatly Many hardware RAID systems ensure that a single point of failure will not stop the functioning of the system by using  Redundant power supplies with battery backup  Multiple controllers and multiple interconnections to guard against controller/interconnection failures Database System Concepts - 5th Edition, Aug 12, 2005. 11.32 ©Silberschatz, Korth and Sudarshan Optical Disks    Compact disk-read only memory (CD-ROM)  Removable disks, 640 MB per disk  Seek time about 100 msec (optical read head is heavier and slower)  Higher latency (3000 RPM) and lower data-transfer rates (3-6 MB/s) compared to magnetic disks Digital Video Disk (DVD)  DVD-5 holds 4.7 GB , and DVD-9 holds 8.5 GB  DVD-10 and DVD-18 are double sided formats with capacities of 9.4 GB and 17 GB  Slow seek time, for same reasons as CD-ROM Record once versions (CD-R and DVD-R) are popular  data can only be written once, and cannot be erased.  high capacity and long lifetime; used for archival storage  Multi-write versions (CD-RW, DVD-RW, DVD+RW and DVD-RAM) also available Database System Concepts - 5th Edition, Aug 12, 2005. 11.33 ©Silberschatz, Korth and Sudarshan Magnetic Tapes   Hold large volumes of data and provide high transfer rates  Few GB for DAT (Digital Audio Tape) format, 10-40 GB with DLT (Digital Linear Tape) format, 100 GB+ with Ultrium format, and 330 GB with Ampex helical scan format  Transfer rates from few to 10s of MB/s Currently the cheapest storage medium   Tapes are cheap, but cost of drives is very high Very slow access time in comparison to magnetic disks and optical disks   limited to sequential access. Some formats (Accelis) provide faster seek (10s of seconds) at cost of lower capacity  Used mainly for backup, for storage of infrequently used information, and as an off-line medium for transferring information from one system to another.  Tape jukeboxes used for very large capacity storage  (terabyte (1012 bytes) to petabye (1015 bytes) Database System Concepts - 5th Edition, Aug 12, 2005. 11.34 ©Silberschatz, Korth and Sudarshan Storage Access  A database file is partitioned into fixed-length storage units called blocks. Blocks are units of both storage allocation and data transfer.  Database system seeks to minimize the number of block transfers between the disk and memory. We can reduce the number of disk accesses by keeping as many blocks as possible in main memory.  Buffer – portion of main memory available to store copies of disk blocks.  Buffer manager – subsystem responsible for allocating buffer space in main memory. Database System Concepts - 5th Edition, Aug 12, 2005. 11.35 ©Silberschatz, Korth and Sudarshan Buffer Manager  Programs call on the buffer manager when they need a block from disk. 1. If the block is already in the buffer, buffer manager returns the address of the block in main memory 2. If the block is not in the buffer, the buffer manager 1. Allocates space in the buffer for the block 1. Replacing (throwing out) some other block, if required, to make space for the new block. 2. Replaced block written back to disk only if it was modified since the most recent time that it was written to/fetched from the disk. 2. Reads the block from the disk to the buffer, and returns the address of the block in main memory to requester. Database System Concepts - 5th Edition, Aug 12, 2005. 11.36 ©Silberschatz, Korth and Sudarshan Buffer-Replacement Policies  Most operating systems replace the block least recently used (LRU strategy)  Idea behind LRU – use past pattern of block references as a predictor of future references  Queries have well-defined access patterns (such as sequential scans), and a database system can use the information in a user’s query to predict future references  LRU can be a bad strategy for certain access patterns involving repeated scans of data   e.g. when computing the join of 2 relations r and s by a nested loops for each tuple tr of r do for each tuple ts of s do if the tuples tr and ts match … Mixed strategy with hints on replacement strategy provided by the query optimizer is preferable Database System Concepts - 5th Edition, Aug 12, 2005. 11.37 ©Silberschatz, Korth and Sudarshan Buffer-Replacement Policies (Cont.)  Pinned block – memory block that is not allowed to be written back to disk.  Toss-immediate strategy – frees the space occupied by a block as soon as the final tuple of that block has been processed  Most recently used (MRU) strategy – system must pin the block currently being processed. After the final tuple of that block has been processed, the block is unpinned, and it becomes the most recently used block.  Buffer manager can use statistical information regarding the probability that a request will reference a particular relation   E.g., the data dictionary is frequently accessed. Heuristic: keep datadictionary blocks in main memory buffer Buffer managers also support forced output of blocks for the purpose of recovery (more in Chapter 17) Database System Concepts - 5th Edition, Aug 12, 2005. 11.38 ©Silberschatz, Korth and Sudarshan File Organization  The database is stored as a collection of files. Each file is a sequence of records. A record is a sequence of fields.  One approach:  assume record size is fixed  each file has records of one particular type only  different files are used for different relations This case is easiest to implement; will consider variable length records later. Database System Concepts - 5th Edition, Aug 12, 2005. 11.39 ©Silberschatz, Korth and Sudarshan Fixed-Length Records  Simple approach:  Store record i starting from byte n  (i – 1), where n is the size of each record.  Record access is simple but records may cross blocks   Modification: do not allow records to cross block boundaries Deletion of record I: alternatives:  move records i + 1, . . ., n to i, . . . , n – 1  move record n to i  do not move records, but link all free records on a free list Database System Concepts - 5th Edition, Aug 12, 2005. 11.40 ©Silberschatz, Korth and Sudarshan Free Lists  Store the address of the first deleted record in the file header.  Use this first record to store the address of the second deleted record, and so on  Can think of these stored addresses as pointers since they “point” to the location of a record.  More space efficient representation: reuse space for normal attributes of free records to store pointers. (No pointers stored in in-use records.) Database System Concepts - 5th Edition, Aug 12, 2005. 11.41 ©Silberschatz, Korth and Sudarshan Variable-Length Records  Variable-length records arise in database systems in several ways:  Storage of multiple record types in a file.  Record types that allow variable lengths for one or more fields.  Record types that allow repeating fields (used in some older data models). Database System Concepts - 5th Edition, Aug 12, 2005. 11.42 ©Silberschatz, Korth and Sudarshan Variable-Length Records: Slotted Page Structure  Slotted page header contains:  number of record entries  end of free space in the block  location and size of each record  Records can be moved around within a page to keep them contiguous with no empty space between them; entry in the header must be updated.  Pointers should not point directly to record — instead they should point to the entry for the record in header. Database System Concepts - 5th Edition, Aug 12, 2005. 11.43 ©Silberschatz, Korth and Sudarshan Organization of Records in Files  Heap – a record can be placed anywhere in the file where there is space  Sequential – store records in sequential order, based on the value of the search key of each record  Hashing – a hash function computed on some attribute of each record; the result specifies in which block of the file the record should be placed  Records of each relation may be stored in a separate file. In a multitable clustering file organization records of several different relations can be stored in the same file  Motivation: store related records on the same block to minimize I/O Database System Concepts - 5th Edition, Aug 12, 2005. 11.44 ©Silberschatz, Korth and Sudarshan Sequential File Organization  Suitable for applications that require sequential processing of the entire file  The records in the file are ordered by a search-key Database System Concepts - 5th Edition, Aug 12, 2005. 11.45 ©Silberschatz, Korth and Sudarshan Sequential File Organization (Cont.)  Deletion – use pointer chains  Insertion –locate the position where the record is to be inserted   if there is free space insert there  if no free space, insert the record in an overflow block  In either case, pointer chain must be updated Need to reorganize the file from time to time to restore sequential order Database System Concepts - 5th Edition, Aug 12, 2005. 11.46 ©Silberschatz, Korth and Sudarshan Multitable Clustering File Organization Store several relations in one file using a multitable clustering file organization Database System Concepts - 5th Edition, Aug 12, 2005. 11.47 ©Silberschatz, Korth and Sudarshan Multitable Clustering File Organization (cont.) Multitable clustering organization of customer and depositor:  good for queries involving depositor single customer and his accounts  bad for queries involving only customer  results in variable size records  Can add pointer chains to link records of a particular relation Database System Concepts - 5th Edition, Aug 12, 2005. customer, and for queries involving one 11.48 ©Silberschatz, Korth and Sudarshan Data Dictionary Storage Data dictionary (also called system catalog) stores metadata: that is, data about data, such as  Information about relations  names of relations  names and types of attributes of each relation  names and definitions of views  integrity constraints  User and accounting information, including passwords  Statistical and descriptive data    number of tuples in each relation Physical file organization information  How relation is stored (sequential/hash/…)  Physical location of relation Information about indices (Chapter 12) Database System Concepts - 5th Edition, Aug 12, 2005. 11.49 ©Silberschatz, Korth and Sudarshan Data Dictionary Storage (Cont.)   Catalog structure  Relational representation on disk  specialized data structures designed for efficient access, in memory A possible catalog representation: Relation_metadata = (relation_name, number_of_attributes, storage_organization, location ) Attribute_metadata = (attribute_name, relation_name, domain_type, position, length) User_metadata = (user_name, encrypted_password, group) Index_metadata = (index_name, relation_name, index_type, index_attributes) View_metadata = (view_name, definition) Database System Concepts - 5th Edition, Aug 12, 2005. 11.50 ©Silberschatz, Korth and Sudarshan End of Chapter Database System Concepts, 5th Ed. ©Silberschatz, Korth and Sudarshan See www.db-book.com for conditions on re-use Record Representation  Records with fixed length fields are easy to represent  Similar to records (structs) in programming languages  Extensions to represent null values   E.g. a bitmap indicating which attributes are null Variable length fields can be represented by a pair (offset,length) where offset is the location within the record and length is field length.  All fields start at predefined location, but extra indirection required for variable length fields A-102 account_number 10 400 Perryridge balance branch_name Example record structure of account record Database System Concepts - 5th Edition, Aug 12, 2005. 11.52 ©Silberschatz, Korth and Sudarshan File Containing account Records Database System Concepts - 5th Edition, Aug 12, 2005. 11.53 ©Silberschatz, Korth and Sudarshan File of Figure 11.6, with Record 2 Deleted and All Records Moved Database System Concepts - 5th Edition, Aug 12, 2005. 11.54 ©Silberschatz, Korth and Sudarshan File of Figure 11.6, With Record 2 deleted and Final Record Moved Database System Concepts - 5th Edition, Aug 12, 2005. 11.55 ©Silberschatz, Korth and Sudarshan Byte-String Representation of Variable-Length Records Database System Concepts - 5th Edition, Aug 12, 2005. 11.56 ©Silberschatz, Korth and Sudarshan Clustering File Structure Database System Concepts - 5th Edition, Aug 12, 2005. 11.57 ©Silberschatz, Korth and Sudarshan Clustering File Structure With Pointer Chains Database System Concepts - 5th Edition, Aug 12, 2005. 11.58 ©Silberschatz, Korth and Sudarshan The depositor Relation Database System Concepts - 5th Edition, Aug 12, 2005. 11.59 ©Silberschatz, Korth and Sudarshan The customer Relation Database System Concepts - 5th Edition, Aug 12, 2005. 11.60 ©Silberschatz, Korth and Sudarshan Clustering File Structure Database System Concepts - 5th Edition, Aug 12, 2005. 11.61 ©Silberschatz, Korth and Sudarshan Database System Concepts - 5th Edition, Aug 12, 2005. 11.62 ©Silberschatz, Korth and Sudarshan Figure 11.4 Database System Concepts - 5th Edition, Aug 12, 2005. 11.63 ©Silberschatz, Korth and Sudarshan Figure 11.7 Database System Concepts - 5th Edition, Aug 12, 2005. 11.64 ©Silberschatz, Korth and Sudarshan Figure 11.8 Database System Concepts - 5th Edition, Aug 12, 2005. 11.65 ©Silberschatz, Korth and Sudarshan Figure 11.100 Database System Concepts - 5th Edition, Aug 12, 2005. 11.66 ©Silberschatz, Korth and Sudarshan Figure 11.20 Database System Concepts - 5th Edition, Aug 12, 2005. 11.67 ©Silberschatz, Korth and Sudarshan Byte-String Representation of Variable-Length Records Byte string representation Attach an end-of-record () control character to the end of each record Difficulty with deletion Difficulty with growth Database System Concepts - 5th Edition, Aug 12, 2005. 11.68 ©Silberschatz, Korth and Sudarshan Fixed-Length Representation   Use one or more fixed length records:  reserved space  pointers Reserved space – can use fixed-length records of a known maximum length; unused space in shorter records filled with a null or end-of-record symbol. Database System Concepts - 5th Edition, Aug 12, 2005. 11.69 ©Silberschatz, Korth and Sudarshan Pointer Method  Pointer method  A variable-length record is represented by a list of fixed-length records, chained together via pointers.  Can be used even if the maximum record length is not known Database System Concepts - 5th Edition, Aug 12, 2005. 11.70 ©Silberschatz, Korth and Sudarshan Pointer Method (Cont.)  Disadvantage to pointer structure; space is wasted in all records except the first in a a chain.  Solution is to allow two kinds of block in file:  Anchor block – contains the first records of chain  Overflow block – contains records other than those that are the first records of chairs. Database System Concepts - 5th Edition, Aug 12, 2005. 11.71 ©Silberschatz, Korth and Sudarshan Mapping of Objects to Files  Mapping objects to files is similar to mapping tuples to files in a relational system; object data can be stored using file structures.  Objects in O-O databases may lack uniformity and may be very large; such objects have to managed differently from records in a relational system.  Set fields with a small number of elements may be implemented using data structures such as linked lists.  Set fields with a larger number of elements may be implemented as separate relations in the database.  Set fields can also be eliminated at the storage level by normalization.  Similar to conversion of multivalued attributes of E-R diagrams to relations Database System Concepts - 5th Edition, Aug 12, 2005. 11.72 ©Silberschatz, Korth and Sudarshan Mapping of Objects to Files (Cont.)  Objects are identified by an object identifier (OID); the storage system needs a mechanism to locate an object given its OID (this action is called dereferencing).  logical identifiers do not directly specify an object’s physical location; must maintain an index that maps an OID to the object’s actual location.  physical identifiers encode the location of the object so the object can be found directly. Physical OIDs typically have the following parts: 1. a volume or file identifier 2. a page identifier within the volume or file 3. an offset within the page Database System Concepts - 5th Edition, Aug 12, 2005. 11.73 ©Silberschatz, Korth and Sudarshan Management of Persistent Pointers  Physical OIDs may be a unique identifier. This identifier is stored in the object also and is used to detect references via dangling pointers. Database System Concepts - 5th Edition, Aug 12, 2005. 11.74 ©Silberschatz, Korth and Sudarshan Management of Persistent Pointers (Cont.)  Implement persistent pointers using OIDs; persistent pointers are substantially longer than are in-memory pointers  Pointer swizzling cuts down on cost of locating persistent objects already inmemory.  Software swizzling (swizzling on pointer deference)  When a persistent pointer is first dereferenced, the pointer is swizzled (replaced by an in-memory pointer) after the object is located in memory.  Subsequent dereferences of of the same pointer become cheap.  The physical location of an object in memory must not change if swizzled pointers pont to it; the solution is to pin pages in memory  When an object is written back to disk, any swizzled pointers it contains need to be unswizzled. Database System Concepts - 5th Edition, Aug 12, 2005. 11.75 ©Silberschatz, Korth and Sudarshan Hardware Swizzling  With hardware swizzling, persistent pointers in objects need the same amount of space as in-memory pointers — extra storage external to the object is used to store rest of pointer information.  Uses virtual memory translation mechanism to efficiently and transparently convert between persistent pointers and in-memory pointers.  All persistent pointers in a page are swizzled when the page is first read in.  thus programmers have to work with just one type of pointer, i.e., inmemory pointer.  some of the swizzled pointers may point to virtual memory addresses that are currently not allocated any real memory (and do not contain valid data) Database System Concepts - 5th Edition, Aug 12, 2005. 11.76 ©Silberschatz, Korth and Sudarshan Hardware Swizzling  Persistent pointer is conceptually split into two parts: a page identifier, and an offset within the page.  The page identifier in a pointer is a short indirect pointer: Each page has a translation table that provides a mapping from the short page identifiers to full database page identifiers.  Translation table for a page is small (at most 1024 pointers in a 4096 byte page with 4 byte pointer)  Multiple pointers in page to the same page share same entry in the translation table. Database System Concepts - 5th Edition, Aug 12, 2005. 11.77 ©Silberschatz, Korth and Sudarshan Hardware Swizzling (Cont.)  Page image before swizzling (page located on disk) Database System Concepts - 5th Edition, Aug 12, 2005. 11.78 ©Silberschatz, Korth and Sudarshan Hardware Swizzling (Cont.)  When system loads a page into memory the persistent pointers in the page are swizzled as described below 1. Persistent pointers in each object in the page are located using object type information 2. For each persistent pointer (pi, oi) find its full page ID Pi 1. If Pi does not already have a virtual memory page allocated to it, allocate a virtual memory page to Pi and read-protect the page 3.  Note: there need not be any physical space (whether in memory or on disk swap-space) allocated for the virtual memory page at this point. Space can be allocated later if (and when) Pi is accessed. In this case read-protection is not required.  Accessing a memory location in the page in the will result in a segmentation violation, which is handled as described later 2. Let vi be the virtual page allocated to Pi (either earlier or above) 3. Replace (pi, oi) by (vi, oi) Replace each entry (pi, Pi) in the translation table, by (vi, Pi) Database System Concepts - 5th Edition, Aug 12, 2005. 11.79 ©Silberschatz, Korth and Sudarshan Hardware Swizzling (Cont.)  When an in-memory pointer is dereferenced, if the operating system detects the page it points to has not yet been allocated storage, or is read-protected, a segmentation violation occurs.  The mmap() call in Unix is used to specify a function to be invoked on segmentation violation  The function does the following when it is invoked 1. Allocate storage (swap-space) for the page containing the referenced address, if storage has not been allocated earlier. Turn off readprotection 2. Read in the page from disk 3. Perform pointer swizzling for each persistent pointer in the page, as described earlier Database System Concepts - 5th Edition, Aug 12, 2005. 11.80 ©Silberschatz, Korth and Sudarshan Hardware Swizzling (Cont.) Page image after swizzling  Page with short page identifier 2395 was allocated address 5001. Observe change in pointers and translation table.  Page with short page identifier 4867 has been allocated address 4867. No change in pointer and translation table. Database System Concepts - 5th Edition, Aug 12, 2005. 11.81 ©Silberschatz, Korth and Sudarshan Hardware Swizzling (Cont.)  After swizzling, all short page identifiers point to virtual memory addresses allocated for the corresponding pages  functions accessing the objects are not even aware that it has persistent pointers, and do not need to be changed in any way!  can reuse existing code and libraries that use in-memory pointers  After this, the pointer dereference that triggered the swizzling can continue  Optimizations:   If all pages are allocated the same address as in the short page identifier, no changes required in the page!  No need for deswizzling — swizzled page can be saved as-is to disk  A set of pages (segment) can share one translation table. Pages can still be swizzled as and when fetched (old copy of translation table is needed). A process should not access more pages than size of virtual memory — reuse of virtual memory addresses for other pages is expensive Database System Concepts - 5th Edition, Aug 12, 2005. 11.82 ©Silberschatz, Korth and Sudarshan Disk versus Memory Structure of Objects  The format in which objects are stored in memory may be different from the formal in which they are stored on disk in the database. Reasons are:  software swizzling – structure of persistent and in-memory pointers are different  database accessible from different machines, with different data representations  Make the physical representation of objects in the database independent of the machine and the compiler.  Can transparently convert from disk representation to form required on the specific machine, language, and compiler, when the object (or page) is brought into memory. Database System Concepts - 5th Edition, Aug 12, 2005. 11.83 ©Silberschatz, Korth and Sudarshan Large Objects  Large objects : binary large objects (blobs) and character large objects (clobs)   Examples include:  text documents  graphical data such as images and computer aided designs audio and video data Large objects may need to be stored in a contiguous sequence of bytes when brought into memory.  If an object is bigger than a page, contiguous pages of the buffer pool must be allocated to store it.  May be preferable to disallow direct access to data, and only allow access through a file-system-like API, to remove need for contiguous storage. Database System Concepts - 5th Edition, Aug 12, 2005. 11.84 ©Silberschatz, Korth and Sudarshan Modifying Large Objects   If the application requires insert/delete of bytes from specified regions of an object:  B+-tree file organization (described later in Chapter 12) can be modified to represent large objects  Each leaf page of the tree stores between half and 1 page worth of data from the object Special-purpose application programs outside the database are used to manipulate large objects:  Text data treated as a byte string manipulated by editors and formatters.  Graphical data and audio/video data is typically created and displayed by separate application  checkout/checkin method for concurrency control and creation of versions Database System Concepts - 5th Edition, Aug 12, 2005. 11.85 ©Silberschatz, Korth and Sudarshan