Chapter 13: Query Processing

صفحه 1:
Chopter 09: Guevy Provessiag 

صفحه 2:
wo Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 

صفحه 3:
عجوو عووسرة1) برست) ۱۰ جم9ظ) مزوه) Telational algebra expression execution plan parser and translator ). @arsicn, ced trocetatics ©. Optcrtzctics: Ocsdrer Gyre Oncewp -O* Crim, Oxi ?, OOOO. 

صفحه 4:
ما ‎vad‏ و ۲ © troehte the query iv ty toteroal Porc. Dhis tp thea troushited tnt rekticerd ‏و‎ © Parser checks syotex, ‏ات او‎ ۲ ‏مسر‎ © Dhe qen-exevuivd coger tohes 0 quer -evotudiog phi, exevuier trot ptr, wood returas the ‏مكلا صا صصص‎ query. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. we ©Sbervehnts, Cork ced Cnakershe 

صفحه 5:
+ ®wir Gteps to Query Provessiay ! Opicoizaiva وه موه روت سا رو موه ماو ‎BO rekticad‏ ص ‎eet‏ (()ی‌یرآآموم‌مسسیر بت ۶ (( )دونو يي" )سن [آ مسموكاناك امصصد خاه صمج عصاص لاله سا موه من ماود مت و ۲ دمد سره رومت بحيب و لاه با نوی یماسا ناسون سره ‎om‏ لس و وه مه الط زرط و له( ۳ ‎Cox, con ee on ike oo bokrare tp Prd coconts wik boknee < CS0O,‏ © © pee peforw owl relia sro ood discard umeructs wits bobo = esoo Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wo ©Sbervehnts, Cork ced Cnakershe 

صفحه 6:
@wic Cteps: Opicotzciva (Ovd.) ۲ Query Opkotraton: @omnpt df equidedt evckniiva phos chover the oor wit: ‏اهاط‎ © Costis esitvoted ustay stisticd iPoreatica Proc the ‏ای ماو‎ ۱3 tuples fo eur reticg, size oP tuples, ete. 4a this chopter we sty © ‏ی وا تسا‎ query ‏و‎ ‎۶ ‏عون ولو اما ماه ۳ لس‎ ۰ اه وا و وه الط ۳ لو او و تیور موه موی 062 سس 1۰ 18 اس ام مه مه لو ‎queries, thot is, hous to‏ وه و تما ‎Oe int‏ © اوه امه مسا Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wo ©Sbervehnts, Cork ced Cnakershe 

صفحه 7:
و + © Costs ‏تسه تسج رای‎ elapsed tere Por oeewerteny query © ‏و۳ نوت‎ contribute to thre ost ° deh urresers, OPO, or even vetwork cow micuticr ۱ she predocinnt cost, ond doy reltivel easy wesc. ‏امه اذا يكام برط لمجت(‎ ۶ Durvber of seeks ‏سس سم ها‎ © ‏وا چاه لین(‎ reed * ‏امه‎ ‎© ‏ای(‎ of blocks writen * averice-blcb-urite-orst ۱ Cost ip varie a block ts ‏اما و ی و نومه ی‎ ~ dot is reed bark oPier betay writes to posure thet the write ‏اوه رد‎ Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wor ©Sbervehnts, Cork ced Cnakershe 

صفحه 8:
+ Deanree of Chery Cost Ora} or skophciy we ist use he outer of block roosters Prow dts oer the ouster of sens us te cost weasures سا ی ات تا سا ‎hy‏ © © py ewe Por ‏مه سس‎ © Oost Por b bck ‏عم سا‎ 6 seeks bin tO *he ۲ Oe keore OPO costs Por steph © Red systews do tohe OPO pps! tate cer uct BE We ue oot eck cost to writen output to disk i 2a post Porat Bl Geverd ckerthws cos reduce deb 10 by wien extra bubPer space © Bonn oF red wewory woluble ty buPPer depeuds va ver pared queries and OG processes, koowe ‏شحو رد بای‎ * Oe Dew use worst cose esikvoies, ‏خأن امه مج عل رای مج‎ ‏رو‎ ueeded Por the pperaiza ts avctkable BE Required data way be beer resideut dread, avoickory dh VO © Cita the ki oct ‏مه اجه بو‎ Ez Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wo ©Sbervehnts, Cork ced Cnakershe 

صفحه 9:
+ Getsvioa Operuiva © te sews — seurck ckpritiws thot bode ord retieve revords thot PUP + ‏ین ماود‎ © Okprikw Od (few search). Groa rack Ale block ocd test ol records to see whether they sutePy the selevio ooodtic. © vet porate = & block iexmPers + (seeks ۱ | ‏سل‎ ounober of blocks evuttciegy records Pow retatica © seevivg & 700 hey otirbute, co sippy oo Preorder, record < oost = (b, 12) block ‏علد 0 + حطس‎ © Liew search en be opted ‏سس‎ of © selevtica ‏عد الم‎ *ardertag oF records te the Pie, or ۲ ‏سل اه امه‎ Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wo ©Sbervehnts, Cork ced Cnakershe 

صفحه 10:
+ Oxbvtra Operctiva (Ova) ۲ 08 (beep seach). Oppiroble P seleviog fy on equdliy cowperisva oo the ‏مرت من هه‎ Pie ts ordered. © ‏سوه وه بو ما هخا ما سا با موه‎ © pet peter (cuncber of dk blocks ty be screed): poet oF ocak he Bist tiple bya biory search on the blocks ‏ابا‎ * (in * to) "AB kere one wnuligls ‏بم بحاصف سسحت‎ seer ~ Ok) roster cost of ‏طساوا اه ای و‎ records that sutsPy selevtiza peedtion: — Ol see how to evttrate this ost to Chapter 1 Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 11:
مسططلت1/ ناس و + لا مهف جوا اس - و بط ۲ .مصلا خن امه من با اه من شلد ۶ ۱ ke correspoentery ‏مه اوه‎ ۶ ‏عبسه‎ 6+0 + 0 BOO (praca index oo wokey, equal) Rerteve walipke records. © Revers ul be on ‏وا مرهج‎ © beth = oxrober oP blocks evoracicny ‏رای‎ records * ‏جدومن‎ * (In the) thot yb BO (equal 7a search-hey oP sevoerckwy krdex). © Retteve ‏كماد ه‎ record Phe searctehey te o rorrdchate hey < ‏جب) ع میسن‎ (۳ (tp + te) © Retteve wutiple records P searck-hey is ont a coackcate hey: ۲ ‏بای‎ of ‏و‎ uichiay records way be oa a dPPereat block * ‏له +6) د و9‎ "(tp * bo) ‏ا‎ E; Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موه‎ 

صفحه 12:
+ ‏اد ی‎ Oonaea eon ۳ ‏اه ماه لت بو‎ the Pore Ope, (1) oF Oy. ofr) by wot © hea Ale soon ‏بصا جه‎ search, © or by ‏له ری‎ fa he Pooky une! I ey ee ere ee) Ree eet oO) < Por Op ‏مه سس من‎ Pied Prot ple > v7 ocd son rekiioa sequel Brow tere > ‏یوت موق‎ )( het soa rekiiva sequeukdy il Prost ‏حصت ان صل إن < جاور‎ keto BO? (sevmnkey index, corre). © Bor ‏ری‎ o(r) vee Rte y Prod Prot iecex eciy > vd sox ‏جلما‎ ‏فسوی‎ Brow here, ty Pind porters 7 records. © Por 69.0 (7) feet soem eck paves of dex ‏جا مر بط‎ Prot ل < بصي ‎١ Aether cose, retteve records that are potted tt‏ ‎requires oa VO Por euck record‏ ~ — bicear Pie soos way be cheuper Eg Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 13:
وت او و دی ی + ۱ i)" yo" = alr) B00 (courte sebvtira ‏مجلم جمد راص‎ © ‏اسلو‎ a obra of 0, cad ckprthors (BC trou (BP that remus far he beet ost Por oy, (1). © Pept cher comics on ple Pier Betcha ht ‏موخطلجا رو‎ ۲ 96 ‏ماش وی مه عسمبس)‎ krber). © Ose ‏رسب‎ cowprste (nutipe-hey) trdex P ovatkble. B00 (corer seb ‏سس روا‎ of kbectPiers). و لس نت یط یی © © Ose correspouday dex Por euch ‏اه اه سول له ,مه‎ vbtdoed sets oP record potaiers. © Dheu Peick records Prow Pie © ۳ ‏ول عون سرت‎ ot hove upproprigie tudes, upply test in wewory. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 14:
و۰ و ‎ers).‏ هل اه مت روا مایت عمط ) 0۵00 ۲ ‎torches.‏ عتالمره جرا میت له خا تلو © ‎hear som.‏ ی مس ( ‎oP oh obraced sets‏ موی لی رمطمت انس وا لبمس تا ‎oP record porters.‏ ‎Dheu Peteh records Brow Pe‏ © 0 مسبت ۲ ام مه ‎Doe brew‏ © ‎OB very Pew reverts satohy 70, cid as keke iF opphoube ty 0‏ ‎tide oad Petck Proc Pier‏ مت سم ماه لب( Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 15:
+ orien © We way bud oo ‏ها من بط‎ rekon, vod feo use the iden t7 reud the rete fo sorted order. (ay feud to vor disk block anvess Por pack tuple. ۲ ‏ما‎ hot Attorney, techoiquer the quicker co be weed. Por ‏ام با متیر‎ Pi ia oer, extorad ‏له سا هبو‎ choice. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 16:
+ @xered Gort Dore Let O dew wewory size (ia power). 4. Orca sorted new. Let tbe 00 ‏.تادهم‎ ‎Reece tr fer Bile ‏:وصتماص جملا خأت عمج جما ألا وجا‎ (6) Grad © Hooke of ‏تس با مق‎ (leer eee tet (6) Orte sorted data ty rust Ri ‏و‎ f bet the Picral value oP tbe D ©. Derg the nme (cent stk)... Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 

صفحه 17:
+ ‏سس‎ OortDerge (Ovu.) ©. Derg te nee (Or were). De wmnve (Por ww) tt D< ۰ (Ove O Docks oP wewory io buPPer iupul ruc, aad ( block to buPPer puiput. Read the Pirst block oP eack rus iio is bubPer pace ۲ ‏امم‎ ‎۱ Ontte the record to the output buPPer. TP the output buPPer is Pull ‏طاسب‎ it to chk. ‎record Prow ts taput bu Per page.‏ با وه( دمل رو و وج بط ‎AB he‏ ‎read the ora block (iP coy) oF te ror fay the bub Per.‏ ‎©. well oop buPPer pages ore expt: ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.07 ©Sbervehnts, Cork ced Cnakershe 

صفحه 18:
+ ‏سس‎ OortDerge (Ovu.) ‎O, severd were pases ar required.‏ 02 الا ‎a eak poss, ooaiknous pups oF OD - nee oe werd.‏ © ‎© © pres ‏امه با لس‎ of rene by a Paci of DA, ord ‏ها مد و‎ by the sae Partor. ‎> Gq. 1۳ ORC, od here ore 80 nes, coe poe redures the ‏ار‎ of nvr i ©, rack UD kos the otze oF the total cer ‎© Repraed passes ore perPorered dl ne Rover brea conreed i ‎pw. ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 19:
14] 19 14 33 هه آج | آع آع 21 ۳ e | 16| AE! ml 3 3 3 16 عماجم outpul merge pass—2 19] 141 33| 311 16 24 14 21 16 [ه ابع] ه] وم دع زه ]ع | ده runs merge 55-1 19 31 24 14 33 16 21 3 16 14 2 d ‏م‎ runs 24 19 31 ¢ [33 b fig e [16 6 d [21 ‎os‏ لبم ‏3 ‏4217 ‏د ‎ ‎ ‎ ‎ ‎initial ‎relation ‎create ‎runs, ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ 

صفحه 20:
+ @xtrad Der Oort (Oow.) 8 ‏ددن‎ os © Pond ‏هچ مه ده امه‎ [lop q(b/D)|- 8) ‏اوه‎ Por intial ruc oreuive os well os ia rack puss iy Ob, > Por Ped poss, we dot count wre cost ‎Por ofl operations sce the mip oP‏ اوه را و جرب اج ی موی ‎he pres‏ مه تا نوی موه ‎los‏ ‏۳ ‏وه اما بوصم سا ان ای اه عط۳؟ ۶ )0 + ا(© لطايمها © ) ا ‎© Geeks! vent side ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 21:
+ @xtrad Der Oort (Oow.) © Oost of seeks © Our, roo ‏ات لو و مه عون تمس‎ ree oed coe geet ‏وا‎ wore wack net «۰ 61۶/۵ © ‏مار موی سنوی‎ < 0 ‏جمد للخصص) رط تساه‎ by, booke of a te) « Weed @[b,/b,] seeks Por euk were puss — except he Prd oe which does oot require Lorie > Dordd ‏میم‎ of seeks! Olt. /O) + ۶/۶ ‏ا(© ما ق)‎ -4( Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 10.60 ©Sbervehnts, Cork ced Cnakershe 

صفحه 22:
00 سس 0 ا + bts Operate © Geverd dPfereat ckerihws ty koplewet joie © Destedtoop ion © Ol cestedtoup pn © Aadexed ‏مام روط لصحي‎ © ‏یی‎ ‎© ‏سیر‎ ‎۲ ‏وت و مس اما ان‎ ۲ ‏ار‎ use the Policy Porvoo ‏اه ین و‎ records oP meter: 10,000 © ‏ی(‎ of blocks oP ‏توص‎ 00 Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏موم‎ 

صفحه 23:
واول جوع( x 0 B Voom he betas or Por ewk tek tard bey Por pack tele ‏م را‎ do boos teot par (1,1) pee B hey sutoPy the fora oxndtirs 0 ‏تج‎ db, old 161 he rem. ead ood Br soled the outer ‏مسق‎ ond 5 the trey rekaion oF the jpic. Qequires oo tedices ced cod be used ut coy bred oF jot ‏.لحم‎ © Cxpewive swe tt exacwtees every pair oP tuples te the tivo rekon. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 

صفحه 24:
Oeste oop \Iota (Ov) he wort exe, Phere by exerts wrexoory aly to hkl ore Hock oP mack retin, bor roma eat ab, +b ‏عام سس اسان‎ we vec AP he miter relatos Pie ‏سوت رو و هی‎ thot ey fe kre rekon 9 ‏2ن عند سيك‎ 2 oveler ‏وج ی او و‎ italy cot motte ‏جا‎ ‎۱ ‎» SOOO * FOO + OO = C,000 400 beck ‏,اس‎ ‎+ SOOO + ADO 2 5000 ‏سس‎ ‎© atk etree oe he enter rektins » OOO * MOO + POO = 1,000 OO bhek ‏سس‎ xed POO verbs AP sxvcler rekitiva (depretor) Pte rated ‏,نوج و‎ the cost ‏مهوت‎ wil be GOO blook ‏امد‎ lock crviexHoue ckprthen (sent okde) by prePrrbke. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. woe ©Sbervehnts, Cork ced Cnakershe 

صفحه 25:
+ Obok Ovsted Loup lot BE Ocrtoct oP cestebop ota fa hick every block of ioc rekon ts poeed ik every block oP ‏ی‎ reson. Por pack block BoP reo bes Por wack beck B, oP ade boots Por cack hk | is @, bo bers Bor pak tele | ta , do bets Obeok B (1,,) 2utePy the joka raredirs Pthey do, kd (he rem ood ood vod vod Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 26:
Obok Ovverdvop ‏بل‎ )00( 19 ‏حلييد ا * © + موخاصمن لجالا را + را + ا مجهي حم جد‎ © Book block ia he tour rektiog ote read core Por pack berks ther outer rekon ‏امس‎ of pare Por euch tiple fr he outer neko: © Qest we! b, + b, beck ixnePers + © secks. 11 eprevewet ty vested loop oad block oested loop ‏زوا‎ ‎© 415 ‏خاصصاط‎ vesieHoop, we D — 2 disk blocks os blochter, wait Por outer rebtiogs, where (= wewory size to blocks; ‏انا ما بو و‎ ‏الا‎ tocar ‏ی ای میس‎ ٠ ‏عا + يعا + |[©-0) / | د سمه‎ beck ‏مج‎ + ‏عد ال©-0) / 1و‎ © 18 ‏عونا جتجاطاد ماصفجج‎ a hey or trey rehaion, sippy tocar bop oo Pirst ‏او‎ © Gow tower bop Porword vad backward ohercately, to wohe ‏حوب‎ oP the blocks rewatoicy it bubPer (wi LRO rephacewect) © Ope ‏لت‎ oo ‏ماس مج‎ P avalible (cert okde) 4 Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 27:
دامل مومرلسوو() وله 19 ‏خا مه با سار مت ما یا‎ © pia te og equtont or oxturd ic ood ‏ری با سوم‎ po the erer rekiiva's parr ‏مجاه‎ ‎» Oem comming on fide et ۵ ‏وروت‎ «kt. BE Por eek ape | she puter rekiion rn, oe he tare took up tikes ts tral sate the ors ord wth tuple. BH Worst cwe: bubPer hos spare Por ool) ove page of 7 oad, Por puck tuple in we perfor os iden ‏بح مب ونیا‎ 9 ‏با اه بسن‎ b (tp th.) + eo © Dhere op he ont of raversien kde ood Petehieg oll wrath, & ple Por oe tuple oer © 2 ‏داوم جد مدو‎ us ovat of a sine selevio oo stay he joka cokion. ۲ 1 ‏ی‎ ore avaluble va jot utributes oP bois road ‏بع‎ ‎we the retaica wih Pewer tuples os he outer retatiza. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wor ©Sbervehnts, Cork ced Cnakershe 

صفحه 28:
ون سل وال سس( ‎Baccople oF‏ + B Ovwnte depres Miter, wil depositor es he vier rekiica. ۱ ‏ما وه را‎ 0 privey (ree fedex oo the ‏بس صصخاصج جفجاطه ماج‎ hick ovotacs OO euttes ta eur terdex ade. © Once nsiwer ke (0,000 tures, the heh oF he ree & P, od vce ore ‏موی بط اما طلست ما مت‎ cea © bret kor SOOO nips Bl Opt of block cented boop ior © OOOO + (OO = POAOO bork ‏سوت‎ + 0 * 00 = COO seeks ) ee ROKR) Worst oe WEY ١ way be stqniPicadly less wi wore weer Oost oP indexed ested ops pric. © dO + SOOO * 9 < 06,100 ‏اه له مامت با‎ © CO ovst thely te be less than thot Por block oested loops ivi Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 29:
ول سوه مور) (0 Gor bok rebiiows on ther ott otebute (Poot dead) sorted oa he pia oirbues). ©. Dene the sorted rektiows ty iota thew (Uo ote7 ts kadar ty the were skp of the sortwerge cherie. Octo dPPereore ts hoodkoy oP duplicate values to pte aide — every pair uit: ‏ای چا اس اه ام دم عنام موی‎ QOetated okprite ia book 4 Orsdrer Gyre Oncewpe -O* Crm, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 30:
(000۰) «اعلسیپ و( ۲ Coa be wed vay Por equijpies oad coturd pice ۲ Guth beck ceeds ty be reed oaiy voce (assur ol tuples Por ‏چاه لس مق بو‎ the ‏ممصم دأ اذا دجاه اط‎ BD be cost oP were ps ‏بط + بط‎ Dock rectors ‏ليطا + ره‎ seeks © + the vost oP sorte P rektions are ‏ی‎ ۲ ‏لوا‎ were pte: IP coe retiiivg is sorted, ued the her hos o sevvodey @4rer todex vo the ‏اه و‎ ‎@*-ree‏ سا اه وت تا ‎Der the sorted rebiion wit the‏ و ‎Gon the result vo the oddresses oP the vested retatiog’s tuples:‏ © ‎tt physivdl address order ood werge wih previo‏ ما موی ما ‎Gown‏ و ‎result, to replace addresses by the ortual tuples‏ ‎۱ Gequectd scoo wore ef Pied tho rondo loobug ‎ ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 همومه‎ 

صفحه 31:
Welsloia BO ppicdbe Por equioics ocd coturd joies. © keek Powis kis used ty portion tuples ‏تا سا اه‎ مه بط و ال و رل ,... ,( ,( ‎ches to‏ جه( )مال سح .لل بصم صمحب مارم ‎oF nocd seed‏ اه © ray ry or deowe portions of rips » Gack te | 6 rie put i portion 1 where 1= hit, [lordPtre]). On Np or dower portions of sheer ١ Bock ‏يا جاص‎ Eee pt in partion &, where 1 = h(i, [lordPire]). Bl Qote: Ie book, ‏جا ب‎ dewted ww “I, 99 sewed we Dont oe ‏يس سوسس‎ سا0 لح 0 لا سواه 1 مهو (ب0 بسط 00 

صفحه 32:
partitions partitions ofr ofs Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. woe ©Sbervehnts, Cork ced Cnakershe 

صفحه 33:
+ Arekxboa (Ova) Br iples tar, ceed oly ty be ‏او‎ wih = hurler ta 5; Deed vt be ‏بود جا ساسج جلاننا لجسو تممه‎ oher portioa, stare! © ‏جاصذك كه لحت طاو مد‎ that soiePy the iota ooadics wll hove the sone ‏اه موز با بو اند‎ |۱۳ kat voke & hushed iy sexoe ude fhe rele hue Wy be ta er ae ‏ب م یو‎ سا0 لح 0 لا سواه 1 موم (ب0 بسط 00 

صفحه 34:
ما2 لب اس + ۳ ‏مایت‎ of ‏.ساد جد لجست جاح اعمس‎ ‏مت با‎ he rektion 9 usr hashing Puurtioa ki Okeo partici, a rekon, ‏جم‎ ‎block of coemory & reserved os the pupal buPPer Ror exch partis. (Portioa rekrtey. 9. or rack ‏بط مسا(‎ oexrory ond bukd oc kecoewory ‏ما مه‎ ‏تاه‎ Dhis hook tex ‏ما موس اس سول وچ‎ ‏اس‎ ‎(b) Read the tuples fa Proc the ‏روا‎ oc. (Por pack tiple |, boca pack ‏ماس‎ pk fir Ain ‏رورس‎ ask krex, Output dor ‏یط و سم سیر‎ (Rektion = ts odled the ‏جنلف‎ ‎te ded the probe top. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. woe ©Sbervehnts, Cork ced Cnakershe 

صفحه 35:
+ Nixebabia deprive (Oot) 5 Dhe uke ood he hook Riccio hy choseu sunk hot eh 9, shoudl Pi is ‏رس‎ ‎© Dypicdly ois chosen oe [b JO] * Phere Pe “Pudge Pastor”, typical won LO © Dhe probe rekon portions 5, ceed ‏معدب جا كا اب‎ ۲ ‏او بجوم موی‎ P ounober of portions ‏اوه ما یی ان‎ ‏وی 0 سوم جر‎ ۶ teal of particcin ou, wee D —( portions Por = © Porter pantion he D =( partiocs weg. dPPerect hark Pucrtoa © ‏لماه عاسم سد دو(‎ oar © Rory required: e.y.,, revursive porticcieg ont ceeded Por rekdioas of (G© or tess wi wewory size oF OOO, wit block size oP PKB. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 666 1 ‏سا0 لح 0 لا سواه‎ 

صفحه 36:
AWaadiay oP OverPlbws © Potticceg is suid to be skewed P sxe portions hove siqaPicadly wore tuples toa sowe vers ۲ ‏رب و و ما و مج رون ابا‎ Rewoocer ‏له‎ © Dow ‏و سای سوه تج نطو‎ otter ۱ ۲ ‏اور نیون‎ com be doce i bull pose © Portion 986 Barker portioned win dPPeredt husk ‏مخ‎ ‎© ‏اه اوه ام ب موه‎ اج وید وی امه و شاه باس تاو وله ‎OverPow‏ © ‎phase‏ © Cy. potion buld rebativg tee wany portions, hea cowbice thea © ok ‏امین ها ات ات ات‎ of ‏امد‎ ‎© albock option: use block vested loops jit vo overPwed portions ka Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 وهم‎ 

صفحه 37:
تا اس ی سس + ۲ 1 ‏جاه بسه تلهم هه را ی‎ kook jpin i> 9+ ۵ ۷ ‏و‎ block tenePers + ©) ‏عاد (أيط/يط| + أيطا/بط|‎ 18 111 ‏و ی‎ required: #۰ ‏ی جه اب‎ required Por ‏مت انا وج‎ ot ‏یروط‎ )( - 0 © best to chovse the swdler rekiiod us he bull ‏وا‎ ‎© Dord vost epiterate ts! OG, + ‏(يط)رريحا رطا‎ - 0[ + &, +b, book ‏وخ‎ + 9) +], ‏اس 1 - ( اما (ارط/‎ OP the ‏هط رتاو و رو و و سا موه نو لا وی‎ required ماخ باج مسح حي طحب ‎Cont‏ © 00 ‏(ب0 بسط‎ wor ©Sbervehnts, Cork ced Cnakershe 

صفحه 38:
+ @xaople oP Ovet ‏دحج اص »شان‎ یط ( جمسسصه ۲ ‏نما میا‎ wewory sie ty OO biochs ‎dOO wd &,, = FOO.‏ = بط ‎depositions to be wed os badd ‏نو‎ Portion t tei Pie portions, euch oP size OO blocks. “Nhs pontizaioy coo be door it zo po. ‎© Giotet), potiicg cusizwer toto Pive porticos,cuck oP size OO. ‏ا عمجل مجه جا جا "!ا"‎ UME puss. ‎18 Vherefore tot ost, quorieg cost oP writer portal Pied blocks: ‎© QO + OO) = (SOO brook texehere + 9) ]6۵0/9[ + ]۵0/9(( = 990 ‏اعد‎ ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 39:
«املحجاعهراا” لجارملا" © OsePd wheo wewory sired wre rehaively forge, ord the bud rout ts bower hort wewory. ۲ Dea Pecture of hybrid hook otc (Keep te Pret partion oF the bud retaiog to wecory. © Cy. Oth wewmy sie oP OS blocks, ‏مه لوط‎ be portioned tia Pave porttives, wack of ste OO blocks. © Oven oP ‏خی‎ ١ Dhe First portion vomupies OO backs oP ‏ی‎ ‎"0 block is used Por iazut, oad ( block eurk Por bubPeriag the viher سس §onstewer's sivtet) pontiioded hic Five porticus ‏اس‎ of size OD © the Pret 5 weed right away Por probing, iesteud of betay write ut © Cost oF O(60 + 980) + CO +00 = (900 bok trocePers Por hybrid bask jot, tosteud oP GOO wil plata hask-jpr. Hybrid kosh-pis wost usePul PD >> le, 1 موص Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 

صفحه 40:
+ 252323038 Bok ‏تم سر د جلاب‎ Mr ‏حاون نووسي‎ © Ciher we ‏عه ,تا ای راما لس‎ © Onwopute he remit oP owe of the skoper jp ro Dd ” Bod result ooeprises those tuples ‏رای ما اف وه ما‎ the recat coerce: One 8 8 gh * 0) Blot wih a divert coaction ‏م‎ ‎00" 990 بت توا ای ناما موی بت ولا ‎oP he records roe‏ رفك ‎os the‏ رون و ‎eo‏ درو ‎seer‏ یی لا لاصو ©) نالع ©) s : x Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wero ©Sbervehnts, Cork ced Cnakershe 

صفحه 41:
+ Oter ‏سم‎ ۲ ‏و را من مووي بجا رجت وا سای‎ sore. © On sort ‏وه لب سود‎ duced to pack cher, od ol but oc set of cherie pet he carat, © Oprrizaicd: duptodtes van be deleted durex) rus yeuerdiva os well os ot ‏بوم امد لت و مات رطس‎ © ALeehin ‏جروت نت مارا - واه سا‎ iio the socve bucket. © ‏تم‎ وا نطو وه دولناد۳ ۶ Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موه‎ 

صفحه 42:
+ 0 ‏م0 عد‎ : Oqyrquiva 0 ‏ود و و و و او‎ ehrmi. اه مسجت سوه سا نا او رما وا لا موه باه ‎Gorter or‏ © .سحب حاصب مج لجتامجت صخا جمدت جمصاصصي ذا صفيصصييعه با مل لو امه مس معط رسد وبحي عمد ججا جا جصارن؟ او سین ۶ جعسامب وروی اور وه نو وه لصو * Cor cow, wor, ‏ای ری با تمه و‎ vo tuples Poucd sv Por ta the yup. ~ Obes cow biciey portal aggregate Por covet, add up the ‏وی‎ ‎٠ Por wy, keep seen ond cant, ood dude suey by count of the cod Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موی‎ 

صفحه 43:
+ ۳ ‏بر‎ ¢ Ost Opercioar Bl Get pperctoe (U, 1 ened —): can ether wor var oP rer yep Per soviet, vrarknd oP keok-p. BE ‏تا وم مهن سا رن‎ ١١ Pantica buts retaiccs vetey he soo hook Branton Proce wack portion fe Powe. 6. ‏د يجو(‎ WPPerect hoki Raccaza, bak! oa reremory husk tender or Proveve 9, <0 Poke © rus Odd hee ‏ره و‎ he hooks ‏خا سلج‎ they are of crear) ‏با للجم بط‎ Apes ta fr hk tke to he reo. ros ۱ ple ta 9:0 the reel ey are cbeeund) there father kets tock ‎Por euch tuple i 9, Pty there ta he hook rice, delete a Proce the‏ با ‎inks. ‎Ot ead oP 9, acd recwatoten fue tn her hash taker ty ther re. ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موی‎ 

صفحه 44:
+ Ober Operdixe 2 BO cter fora cor be ‏اه اس‎ ۶ © jos Poloued by ocklion of ‏موی الوا‎ Kies. © ‏مام صا ور كالح برط‎ keris. ۲ ‏کج ور مت م۵(‎ ۱ tuples ore those tr —T1,(r ‏ل‎ <4 ۶ USD eter eee Pers ery ete Broce ‏ی بت و فا‎ ony tuple ies, mulput | porkded wks cubs. © Right outer pra gad Pull vuerspic coc be computed sila. © Dod jiey hook ‏و صا مام‎ =I OAR he probe rekon, cpu coxmamachen rips porkied wits ‏لج‎ OP re bull rekon, uheo probes heey irak oP hick rte ‏ندج د خم لمج ا( وج نمی‎ sexewricked riuples packted uly cnr Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wee ©Sbervehnts, Cork ced Cnakershe 

صفحه 45:
ار + © Gp Pot we have see ‏شوه امطلی جيك‎ © lersaives Por ‏و باه‎ cutee expressing tree ۶ ‏وی لول‎ resus of on expreseiza whose Rus are rektiocee or are dent) vxuted, utertdie (story) toa deh. Repeat, ‎operctivar eve cer cnt operon te bet‏ مسجو مو لجووج رس سوير يسوم جوم ‎18 We sty cbove diercaives ‏لماع سس دا‎ ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 46:
تن + ۲ ‏ره را وه موه هن ره تمه لبلب(‎ of the fawestlevel. ‏موه بو(‎ results ‏جعصقماه روما ها ای‎ ‏اوه رطف ارو صا‎ وه اجه و ,بنطاصطا متا رو 18 0 accoun ‏ا‎ be Lialepeesnf ACCOU! ce ae te Prppviogs oa Cusine nnn, 1 0 aN | alance < 2500 

صفحه 47:
Qcterintzciva (Ovd.) ۲ ‏ماه لطاب(‎ i chivas uppicdble Cost oF varie resale to db god ready thew buck coo be quite hich: © Our cost Porwukts Por operaives iquore cost of wuritay results to dish, 57 * Overdl cost = Guep oP costs of ‏موه امه‎ + ost oF ‏لح رت‎ results to cet BH Double bubPeriog! use tue vulgut buPPers Por euch operaiva, wheo coe is Pull ‏رید‎ to cists ube the otter ts options Pied © lows overlap oP ‏ول له موی لت رن الط‎ exertion fee Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wer ©Sbervehnts, Cork ced Cnakershe 

صفحه 48:
Pipelsiay © Opeheed euckeiog 2 evoke severd opercivas skoubouevusl, passtay he results oP ‎oa ty the cen.‏ موه عون ‎B 0.x, i previaw expresin tree, dou! store rem of ‏موی سیر‎ ( ‎۱ tuples dreviy to he pts... Grey, doo't store resul oP joi, pass ‎tuples direviy to proevios. ‎© Ouch cheuer thos wateridizdiod! a7 ceed ty store a tewporary rehotivg to deh. ‏اس ‎owt chivas be possible — ex, sort,‏ وی دادم ‎© Cor pipelctay to be ef Pecive, use evolvaivg doris thot yeceruie culgul tuples eve ve tuples ore reveived Por iapuis to the operatic. ‎© pehes coo be exerted tote weve: dewoad dived ued produzer even ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 

صفحه 49:
+ Prpetcicy (Ova) ۳ 1 ‏تم‎ ted or kay eckson © systeww repedtedyy requests vent tuple Prow top level operctios © Cock operciog requests ued tuple Prow chidres ppercioes us required, ist ‏جد جلا ددج صا عولمج‎ tule © Aa betwern cls, ppercica kes ‏وسنصكا 3 مد "اماك " جماماميه صا‎ bol ‏اوه مج صا‎ © (a prodwersdted or exer pipet © Openers produce tuples eagerly ond poss thew up ty ther poreuis: * OP Re, watctatced betwero operdors, chi puts tuples to buPPer, ‏جم‎ ‎rewoves tuples Prow buPPer "OP buPPer ts Pal, child uvatte til here te space to the buPPer, ood theo ‏وس‎ wore tuples ‎god mas process:‏ له ‎pperctioes thot hove spore tt pupal‏ لاه مسق و ‎swore fag tuples‏ ‎Bl of pipet ‎Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. wee ©Sbervehnts, Cork ced Cnakershe 

صفحه 50:
+ Prpetcicy (Ova) © ope wedtatioa of deworddeved pipetsiay © Cock operation is toplewedied os on teruier tepleweitery the Polo tcry ‏رن‎ » ‏سب‎ ‎~ @.y. Pe scorn: ioitalze Pile sot > state! potater to beqcrciccy oP Pie - ‏او و نم و‎ « ‏شاه اوه که مسا وا ور تاد‎ a) - © ox. Por Fle score: Output ced tuple, cod odvoaze ood store Pile: potter ~ Cox, Por werye pit ‏اه او مت ی طلست سا‎ i went cult tuple is Pound. Gave potters us tieruior state. » chose) Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏سا0 لح 0 لا سواه 1 همومه‎ 

صفحه 51:
+ ‏سب ملس‎ Pyetctag ۲ ‏ول وت‎ ore 0 ob ty cup results eves op hey yet fap tuples © Bx. were pin, or bark joa ‏و‎ reply write ty desk oul ec read barks ۴ ‏پم تا‎ to cpoerde (ot bust swe) ress oa he Ply, a fap hips ore ‏امسر‎ ‎© Cox hybrid hook pc sever cies ovina ples eve oe probe ‏اوه مایت‎ he ‏ماسم تدمج يمسر‎ (yuri (D) are rend ts © Pipetced pic teohaique: Wybrid hock jot, wodPied to buPPer portion D tuples of ‏ولو روت مشاه سا‎ thew us they beoowe ovciible, ord culzut resus oP coy watches betwerd portion (D tuples > Okew a caw ry rk Pound, wruch wth enti ‏يد‎ Apes, oul wrtohes, wend ‏و سوت‎ > Gyrowetridy Por 9g) hplew Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. 16.60 ©Sbervehnts, Cork ced Cnakershe 

صفحه 52:
Gad oP Okaper 

صفحه 53:
11 «۵ 9 balance < 2500; use index account Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. ‏موم‎ 

صفحه 54:
عدادل مبعام جو 2) 13 ‏هما تسد صل‎ rektiows! ber ‏لصفو‎ mete Mer 11 Oreew 1. One ue ‏لاصيا‎ ١ ‏لومجاطد‎ wer result io cocopute: ben (cepesior bo) ۲ ‏جب ,اوكا ودبتل ماج رصیق‎ thea iota the results صموسحه © Groew 9. Perform the parr oP oe of are. Build ond totes oat bro Por ‏بخاص عونا -صصصطصت جه لمه جوتأ صم مما‎ © or cack tuple fa deposi, look up he correepoerden tuples tt puter wand he corresponntey tuples ta bro. © Gack tuple oP deposits exaviced exactly voce. © Grrotew 9 cowbiees te operdivas itie vor spectohpurpose ‏ها موه‎ ‏مسا راو لا ماه وم‎ pice oF tue rekatiocs. Ocsdrer Gyre Oncewpe -O* Crim, Ooi ?, OOOO. woe ©Sbervehnts, Cork ced Cnakershe 

Chapter 13: Query Processing Database System Concepts, 5th Ed. ©Silberschatz, Korth and Sudarshan See www.db-book.com for conditions on re-use Chapter 13: Query Processing  Overview  Measures of Query Cost  Selection Operation  Sorting  Join Operation  Other Operations  Evaluation of Expressions Database System Concepts - 5th Edition, Aug 27, 2005. 13.2 ©Silberschatz, Korth and Sudarshan Basic Steps in Query Processing 1. Parsing and translation 2. Optimization 3. Evaluation Database System Concepts - 5th Edition, Aug 27, 2005. 13.3 ©Silberschatz, Korth and Sudarshan Basic Steps in Query Processing (Cont.)   Parsing and translation  translate the query into its internal form. This is then translated into relational algebra.  Parser checks syntax, verifies relations Evaluation  The query-execution engine takes a query-evaluation plan, executes that plan, and returns the answers to the query. Database System Concepts - 5th Edition, Aug 27, 2005. 13.4 ©Silberschatz, Korth and Sudarshan Basic Steps in Query Processing : Optimization  A relational algebra expression may have many equivalent expressions  E.g., balance2500(balance(account)) is equivalent to balance(balance2500(account))  Each relational algebra operation can be evaluated using one of several different algorithms   Correspondingly, a relational-algebra expression can be evaluated in many ways. Annotated expression specifying detailed evaluation strategy is called an evaluation-plan.  E.g., can use an index on balance to find accounts with balance < 2500,  or can perform complete relation scan and discard accounts with balance  2500 Database System Concepts - 5th Edition, Aug 27, 2005. 13.5 ©Silberschatz, Korth and Sudarshan Basic Steps: Optimization (Cont.)  Query Optimization: Amongst all equivalent evaluation plans choose the one with lowest cost.  Cost is estimated using statistical information from the database catalog    e.g. number of tuples in each relation, size of tuples, etc. In this chapter we study  How to measure query costs  Algorithms for evaluating relational algebra operations  How to combine algorithms for individual operations in order to evaluate a complete expression In Chapter 14  We study how to optimize queries, that is, how to find an evaluation plan with lowest estimated cost Database System Concepts - 5th Edition, Aug 27, 2005. 13.6 ©Silberschatz, Korth and Sudarshan Measures of Query Cost  Cost is generally measured as total elapsed time for answering query  Many factors contribute to time cost   disk accesses, CPU, or even network communication Typically disk access is the predominant cost, and is also relatively easy to estimate. Measured by taking into account  Number of seeks * average-seek-cost  Number of blocks read * average-block-read-cost  Number of blocks written * average-block-write-cost  Cost to write a block is greater than cost to read a block – data is read back after being written to ensure that the write was successful Database System Concepts - 5th Edition, Aug 27, 2005. 13.7 ©Silberschatz, Korth and Sudarshan Measures of Query Cost (Cont.)  For simplicity we just use the number of block transfers from disk and the number of seeks as the cost measures  tT – time to transfer one block tS – time for one seek  Cost for b block transfers plus S seeks b * tT + S * tS We ignore CPU costs for simplicity  Real systems do take CPU cost into account We do not include cost to writing output to disk in our cost formulae Several algorithms can reduce disk IO by using extra buffer space  Amount of real memory available to buffer depends on other concurrent queries and OS processes, known only during execution  We often use worst case estimates, assuming only the minimum amount of memory needed for the operation is available Required data may be buffer resident already, avoiding disk I/O  But hard to take into account for cost estimation      Database System Concepts - 5th Edition, Aug 27, 2005. 13.8 ©Silberschatz, Korth and Sudarshan Selection Operation  File scan – search algorithms that locate and retrieve records that fulfill a selection condition.  Algorithm A1 (linear search). Scan each file block and test all records to see whether they satisfy the selection condition.  Cost estimate = br block transfers + 1 seek   If selection is on a key attribute, can stop on finding record   br denotes number of blocks containing records from relation r cost = (br /2) block transfers + 1 seek Linear search can be applied regardless of  selection condition or  ordering of records in the file, or  availability of indices Database System Concepts - 5th Edition, Aug 27, 2005. 13.9 ©Silberschatz, Korth and Sudarshan Selection Operation (Cont.)  A2 (binary search). Applicable if selection is an equality comparison on the attribute on which file is ordered.  Assume that the blocks of a relation are stored contiguously  Cost estimate (number of disk blocks to be scanned):  cost of locating the first tuple by a binary search on the blocks  log2(br) * (tT + tS)  If there are multiple records satisfying selection – Add transfer cost of the number of blocks containing records that satisfy selection condition – Will see how to estimate this cost in Chapter 14 Database System Concepts - 5th Edition, Aug 27, 2005. 13.10 ©Silberschatz, Korth and Sudarshan Selections Using Indices  Index scan – search algorithms that use an index   selection condition must be on search-key of index. A3 (primary index on candidate key, equality). Retrieve a single record that satisfies the corresponding equality condition Cost = (hi + 1) * (tT + tS)  A4 (primary index on nonkey, equality) Retrieve multiple records.  Records will be on consecutive blocks  Let b = number of blocks containing matching records  Cost = hi * (tT + tS) + tS + tT * b  A5 (equality on search-key of secondary index).  Retrieve a single record if the search-key is a candidate key  Cost = (hi + 1) * (tT + tS) Retrieve multiple records if search-key is not a candidate key  each of n matching records may be on a different block    Cost = (hi + n) * (tT + tS) – Can be very expensive! Database System Concepts - 5th Edition, Aug 27, 2005. 13.11 ©Silberschatz, Korth and Sudarshan Selections Involving Comparisons    Can implement selections of the form AV (r) or A  V(r) by using  a linear file scan or binary search,  or by using indices in the following ways: A6 (primary index, comparison). (Relation is sorted on A)  For A  V(r) use index to find first tuple  v and scan relation sequentially from there  For AV (r) just scan relation sequentially till first tuple > v; do not use index A7 (secondary index, comparison).  For A  V(r) use index to find first index entry  v and scan index sequentially from there, to find pointers to records.  For AV (r) just scan leaf pages of index finding pointers to records, till first entry > v  In either case, retrieve records that are pointed to – requires an I/O for each record – Linear file scan may be cheaper Database System Concepts - 5th Edition, Aug 27, 2005. 13.12 ©Silberschatz, Korth and Sudarshan Implementation of Complex Selections  Conjunction: 1 2. . . n(r)  A8 (conjunctive selection using one index).    Test other conditions on tuple after fetching it into memory buffer. A9 (conjunctive selection using multiple-key index ).   Select a combination of i and algorithms A1 through A7 that results in the least cost for i (r). Use appropriate composite (multiple-key) index if available. A10 (conjunctive selection by intersection of identifiers).  Requires indices with record pointers.  Use corresponding index for each condition, and take intersection of all the obtained sets of record pointers.  Then fetch records from file  If some conditions do not have appropriate indices, apply test in memory. Database System Concepts - 5th Edition, Aug 27, 2005. 13.13 ©Silberschatz, Korth and Sudarshan Algorithms for Complex Selections  Disjunction:1 2 . . . n (r).  A11 (disjunctive selection by union of identifiers).  Applicable if all conditions have available indices.   Otherwise use linear scan.  Use corresponding index for each condition, and take union of all the obtained sets of record pointers.  Then fetch records from file Negation: (r)  Use linear scan on file  If very few records satisfy , and an index is applicable to   Find satisfying records using index and fetch from file Database System Concepts - 5th Edition, Aug 27, 2005. 13.14 ©Silberschatz, Korth and Sudarshan Sorting  We may build an index on the relation, and then use the index to read the relation in sorted order. May lead to one disk block access for each tuple.  For relations that fit in memory, techniques like quicksort can be used. For relations that don’t fit in memory, external sort-merge is a good choice. Database System Concepts - 5th Edition, Aug 27, 2005. 13.15 ©Silberschatz, Korth and Sudarshan External Sort-Merge Let M denote memory size (in pages). 1. Create sorted runs. Let i be 0 initially. Repeatedly do the following till the end of the relation: (a) Read M blocks of relation into memory (b) Sort the in-memory blocks (c) Write sorted data to run Ri; increment i. Let the final value of i be N 2. Merge the runs (next slide)….. Database System Concepts - 5th Edition, Aug 27, 2005. 13.16 ©Silberschatz, Korth and Sudarshan External Sort-Merge (Cont.) 2. Merge the runs (N-way merge). We assume (for now) that N < M. 1. Use N blocks of memory to buffer input runs, and 1 block to buffer output. Read the first block of each run into its buffer page 2. repeat 3. 1. Select the first record (in sort order) among all buffer pages 2. Write the record to the output buffer. If the output buffer is full write it to disk. 3. Delete the record from its input buffer page. If the buffer page becomes empty then read the next block (if any) of the run into the buffer. until all input buffer pages are empty: Database System Concepts - 5th Edition, Aug 27, 2005. 13.17 ©Silberschatz, Korth and Sudarshan External Sort-Merge (Cont.)  If N  M, several merge passes are required.  In each pass, contiguous groups of M - 1 runs are merged.  A pass reduces the number of runs by a factor of M -1, and creates runs longer by the same factor.   E.g. If M=11, and there are 90 runs, one pass reduces the number of runs to 9, each 10 times the size of the initial runs Repeated passes are performed till all runs have been merged into one. Database System Concepts - 5th Edition, Aug 27, 2005. 13.18 ©Silberschatz, Korth and Sudarshan Example: External Sorting Using Sort-Merge Database System Concepts - 5th Edition, Aug 27, 2005. 13.19 ©Silberschatz, Korth and Sudarshan External Merge Sort (Cont.)  Cost analysis:  Total number of merge passes required: logM–1(br/M).  Block transfers for initial run creation as well as in each pass is 2br  for final pass, we don’t count write cost – we ignore final write cost for all operations since the output of an operation may be sent to the parent operation without being written to disk   Thus total number of block transfers for external sorting: br ( 2 logM–1(br / M) + 1) Seeks: next slide Database System Concepts - 5th Edition, Aug 27, 2005. 13.20 ©Silberschatz, Korth and Sudarshan External Merge Sort (Cont.)  Cost of seeks  During run generation: one seek to read each run and one seek to write each run   2 br / M During the merge phase  Buffer size: bb (read/write bb blocks at a time)  Need 2 br / bb seeks for each merge pass – except the final one which does not require a write  Total number of seeks: 2 br / M + br / bb (2 logM–1(br / M) -1) Database System Concepts - 5th Edition, Aug 27, 2005. 13.21 ©Silberschatz, Korth and Sudarshan Join Operation  Several different algorithms to implement joins  Nested-loop join  Block nested-loop join  Indexed nested-loop join  Merge-join  Hash-join  Choice based on cost estimate  Examples use the following information  Number of records of customer: 10,000  Number of blocks of customer: Database System Concepts - 5th Edition, Aug 27, 2005. 13.22 400 depositor: 5000 depositor: 100 ©Silberschatz, Korth and Sudarshan Nested-Loop Join  To compute the theta join r for each tuple tr in r do begin  s for each tuple ts in s do begin test pair (tr,ts) to see if they satisfy the join condition  if they do, add tr • ts to the result. end end  r is called the outer relation and s the inner relation of the join.  Requires no indices and can be used with any kind of join condition.  Expensive since it examines every pair of tuples in the two relations. Database System Concepts - 5th Edition, Aug 27, 2005. 13.23 ©Silberschatz, Korth and Sudarshan Nested-Loop Join (Cont.)  In the worst case, if there is enough memory only to hold one block of each relation, the estimated cost is n r  bs + b r block transfers, plus n r + br seeks  If the smaller relation fits entirely in memory, use that as the inner relation.   Reduces cost to br + bs block transfers and 2 seeks Assuming worst case memory availability cost estimate is   with depositor as outer relation:  5000  400 + 100 = 2,000,100 block transfers,  5000 + 100 = 5100 seeks with customer as the outer relation  10000  100 + 400 = 1,000,400 block transfers and 10,400 seeks  If smaller relation (depositor) fits entirely in memory, the cost estimate will be 500 block transfers.  Block nested-loops algorithm (next slide) is preferable. Database System Concepts - 5th Edition, Aug 27, 2005. 13.24 ©Silberschatz, Korth and Sudarshan Block Nested-Loop Join  Variant of nested-loop join in which every block of inner relation is paired with every block of outer relation. for each block Br of r do begin for each block Bs of s do begin for each tuple tr in Br do begin for each tuple ts in Bs do begin Check if (tr,ts) satisfy the join condition if they do, add tr • ts to the result. end end end end Database System Concepts - 5th Edition, Aug 27, 2005. 13.25 ©Silberschatz, Korth and Sudarshan Block Nested-Loop Join (Cont.)  Worst case estimate: br  bs + br block transfers + 2 * br seeks  Each block in the inner relation s is read once for each block in the outer relation (instead of once for each tuple in the outer relation  Best case: br + bs block transfers + 2 seeks.  Improvements to nested loop and block nested loop algorithms:  In block nested-loop, use M — 2 disk blocks as blocking unit for outer relations, where M = memory size in blocks; use remaining two blocks to buffer inner relation and output  Cost = br / (M-2)  bs + br block transfers + 2 br / (M-2) seeks  If equi-join attribute forms a key or inner relation, stop inner loop on first match  Scan inner loop forward and backward alternately, to make use of the blocks remaining in buffer (with LRU replacement)  Use index on inner relation if available (next slide) Database System Concepts - 5th Edition, Aug 27, 2005. 13.26 ©Silberschatz, Korth and Sudarshan Indexed Nested-Loop Join  Index lookups can replace file scans if  join is an equi-join or natural join and  an index is available on the inner relation’s join attribute  Can construct an index just to compute a join.  For each tuple tr in the outer relation r, use the index to look up tuples in s that satisfy the join condition with tuple tr.  Worst case: buffer has space for only one page of r, and, for each tuple in r, we perform an index lookup on s.  Cost of the join: br (tT + tS) + nr  c   Where c is the cost of traversing index and fetching all matching s tuples for one tuple or r  c can be estimated as cost of a single selection on s using the join condition. If indices are available on join attributes of both r and s, use the relation with fewer tuples as the outer relation. Database System Concepts - 5th Edition, Aug 27, 2005. 13.27 ©Silberschatz, Korth and Sudarshan Example of Nested-Loop Join Costs  Compute depositor  Let customer have a primary B+-tree index on the join attribute customer-name, which contains 20 entries in each index node.  Since customer has 10,000 tuples, the height of the tree is 4, and one more access is needed to find the actual data  depositor has 5000 tuples  Cost of block nested loops join  400*100 + 100 = 40,100 block transfers + 2 * 100 = 200 seeks    customer, with depositor as the outer relation. assuming worst case memory may be significantly less with more memory Cost of indexed nested loops join  100 + 5000 * 5 = 25,100 block transfers and seeks.  CPU cost likely to be less than that for block nested loops join Database System Concepts - 5th Edition, Aug 27, 2005. 13.28 ©Silberschatz, Korth and Sudarshan Merge-Join 1. Sort both relations on their join attribute (if not already sorted on the join attributes). 2. Merge the sorted relations to join them 1. Join step is similar to the merge stage of the sort-merge algorithm. 2. Main difference is handling of duplicate values in join attribute — every pair with same value on join attribute must be matched 3. Detailed algorithm in book Database System Concepts - 5th Edition, Aug 27, 2005. 13.29 ©Silberschatz, Korth and Sudarshan Merge-Join (Cont.)  Can be used only for equi-joins and natural joins  Each block needs to be read only once (assuming all tuples for any given value of the join attributes fit in memory  Thus the cost of merge join is: br + bs block transfers + br / bb + bs / bb seeks   + the cost of sorting if relations are unsorted. hybrid merge-join: If one relation is sorted, and the other has a secondary B +-tree index on the join attribute  Merge the sorted relation with the leaf entries of the B+-tree .  Sort the result on the addresses of the unsorted relation’s tuples  Scan the unsorted relation in physical address order and merge with previous result, to replace addresses by the actual tuples  Sequential scan more efficient than random lookup Database System Concepts - 5th Edition, Aug 27, 2005. 13.30 ©Silberschatz, Korth and Sudarshan Hash-Join  Applicable for equi-joins and natural joins.  A hash function h is used to partition tuples of both relations  h maps JoinAttrs values to {0, 1, ..., n}, where JoinAttrs denotes the common attributes of r and s used in the natural join.  r0, r1, . . ., rn denote partitions of r tuples   r0,, r1. . ., rn denotes partitions of s tuples   Each tuple tr  r is put in partition ri where i = h(tr [JoinAttrs]). Each tuple ts s is put in partition si, where i = h(ts [JoinAttrs]). Note: In book, ri is denoted as Hri, si is denoted as Hsi and n is denoted as nh. Database System Concepts - 5th Edition, Aug 27, 2005. 13.31 ©Silberschatz, Korth and Sudarshan Hash-Join (Cont.) Database System Concepts - 5th Edition, Aug 27, 2005. 13.32 ©Silberschatz, Korth and Sudarshan Hash-Join (Cont.)  r tuples in ri need only to be compared with s tuples in si Need not be compared with s tuples in any other partition, since:  an r tuple and an s tuple that satisfy the join condition will have the same value for the join attributes.  If that value is hashed to some value i, the r tuple has to be in ri and the s tuple in si. Database System Concepts - 5th Edition, Aug 27, 2005. 13.33 ©Silberschatz, Korth and Sudarshan Hash-Join Algorithm The hash-join of r and s is computed as follows. 1. Partition the relation s using hashing function h. When partitioning a relation, one block of memory is reserved as the output buffer for each partition. 2. Partition r similarly. 3. For each i: (a) Load si into memory and build an in-memory hash index on it using the join attribute. This hash index uses a different hash function than the earlier one h. (b) Read the tuples in ri from the disk one by one. For each tuple tr locate each matching tuple ts in si using the in-memory hash index. Output the concatenation of their attributes. Relation s is called the build input and r is called the probe input. Database System Concepts - 5th Edition, Aug 27, 2005. 13.34 ©Silberschatz, Korth and Sudarshan Hash-Join algorithm (Cont.)   The value n and the hash function h is chosen such that each si should fit in memory.  Typically n is chosen as bs/M * f where f is a “fudge factor”, typically around 1.2  The probe relation partitions si need not fit in memory Recursive partitioning required if number of partitions n is greater than number of pages M of memory.  instead of partitioning n ways, use M – 1 partitions for s  Further partition the M – 1 partitions using a different hash function  Use same partitioning method on r  Rarely required: e.g., recursive partitioning not needed for relations of 1GB or less with memory size of 2MB, with block size of 4KB. Database System Concepts - 5th Edition, Aug 27, 2005. 13.35 ©Silberschatz, Korth and Sudarshan Handling of Overflows  Partitioning is said to be skewed if some partitions have significantly more tuples than some others  Hash-table overflow occurs in partition si if si does not fit in memory. Reasons could be    Many tuples in s with same value for join attributes  Bad hash function Overflow resolution can be done in build phase  Partition si is further partitioned using different hash function.  Partition ri must be similarly partitioned. Overflow avoidance performs partitioning carefully to avoid overflows during build phase   E.g. partition build relation into many partitions, then combine them Both approaches fail with large numbers of duplicates  Fallback option: use block nested loops join on overflowed partitions Database System Concepts - 5th Edition, Aug 27, 2005. 13.36 ©Silberschatz, Korth and Sudarshan Cost of Hash-Join  If recursive partitioning is not required: cost of hash join is 3(br + bs) +4  nh block transfers + 2( br / bb + bs / bb) seeks   If recursive partitioning required:  number of passes required for partitioning build relation s is logM–1(bs) – 1  best to choose the smaller relation as the build relation.  Total cost estimate is: 2(br + bs logM–1(bs) – 1 + br + bs block transfers + 2(br / bb + bs / bb) logM–1(bs) – 1 seeks If the entire build input can be kept in main memory no partitioning is required  Cost estimate goes down to br + bs. Database System Concepts - 5th Edition, Aug 27, 2005. 13.37 ©Silberschatz, Korth and Sudarshan Example of Cost of Hash-Join customer depositor  Assume that memory size is 20 blocks  bdepositor= 100 and bcustomer = 400.  depositor is to be used as build input. Partition it into five partitions, each of size 20 blocks. This partitioning can be done in one pass.  Similarly, partition customer into five partitions,each of size 80. This is also done in one pass.  Therefore total cost, ignoring cost of writing partially filled blocks:  3(100 + 400) = 1500 block transfers + 2( 100/3 + 400/3) = 336 seeks Database System Concepts - 5th Edition, Aug 27, 2005. 13.38 ©Silberschatz, Korth and Sudarshan Hybrid Hash–Join  Useful when memory sized are relatively large, and the build input is bigger than memory.  Main feature of hybrid hash join: Keep the first partition of the build relation in memory.  E.g. With memory size of 25 blocks, depositor can be partitioned into five partitions, each of size 20 blocks.   Division of memory:  The first partition occupies 20 blocks of memory  1 block is used for input, and 1 block each for buffering the other 4 partitions. customer is similarly partitioned into five partitions each of size 80  the first is used right away for probing, instead of being written out Cost of 3(80 + 320) + 20 +80 = 1300 block transfers for hybrid hash join, instead of 1500 with plain hash-join.  Hybrid hash-join most useful if M >> bs  Database System Concepts - 5th Edition, Aug 27, 2005. 13.39 ©Silberschatz, Korth and Sudarshan Complex Joins  Join with a conjunctive condition: r 1  2...   n s  Either use nested loops/block nested loops, or  Compute the result of one of the simpler joins r  i s final result comprises those tuples in the intermediate result that satisfy the remaining conditions 1  . . .  i –1  i +1  . . .  n  Join with a disjunctive condition r   1  2 ...  n s Either use nested loops/block nested loops, or Compute as the union of the records in individual joins r (r 1 Database System Concepts - 5th Edition, Aug 27, 2005. s )  (r 2 s)  . . .  (r 13.40 n  i s: s) ©Silberschatz, Korth and Sudarshan Other Operations   Duplicate elimination can be implemented via hashing or sorting.  On sorting duplicates will come adjacent to each other, and all but one set of duplicates can be deleted.  Optimization: duplicates can be deleted during run generation as well as at intermediate merge steps in external sort-merge.  Hashing is similar – duplicates will come into the same bucket. Projection:  perform projection on each tuple  followed by duplicate elimination. Database System Concepts - 5th Edition, Aug 27, 2005. 13.41 ©Silberschatz, Korth and Sudarshan Other Operations : Aggregation  Aggregation can be implemented in a manner similar to duplicate elimination.  Sorting or hashing can be used to bring tuples in the same group together, and then the aggregate functions can be applied on each group.  Optimization: combine tuples in the same group during run generation and intermediate merges, by computing partial aggregate values  For count, min, max, sum: keep aggregate values on tuples found so far in the group. – When combining partial aggregate for count, add up the aggregates  For avg, keep sum and count, and divide sum by count at the end Database System Concepts - 5th Edition, Aug 27, 2005. 13.42 ©Silberschatz, Korth and Sudarshan Other Operations : Set Operations   Set operations (,  and ): can either use variant of merge-join after sorting, or variant of hash-join. E.g., Set operations using hashing: 1. Partition both relations using the same hash function 2. Process each partition i as follows. 1. Using a different hashing function, build an in-memory hash index on ri. 2. Process si as follows  r  s: 1. Add tuples in si to the hash index if they are not already in it. At end of si add the tuples in the hash index to the result.  r  s: 2. output tuples in si to the result if they are already there in the hash index  r – s: 1. for each tuple in si, if it is there in the hash index, delete it from the index. 2. At end of si add remaining tuples in the hash index to the result. 1. Database System Concepts - 5th Edition, Aug 27, 2005. 13.43 ©Silberschatz, Korth and Sudarshan Other Operations : Outer Join    Outer join can be computed either as  A join followed by addition of null-padded non-participating tuples.  by modifying the join algorithms. Modifying merge join to compute r s  In r s, non participating tuples are those in r – R(r  Modify merge-join to compute r s: During merging, for every tuple tr from r that do not match any tuple in s, output tr padded with nulls.  Right outer-join and full outer-join can be computed similarly. Modifying hash join to compute r s) s  If r is probe relation, output non-matching r tuples padded with nulls  If r is build relation, when probing keep track of which r tuples matched s tuples. At end of si output non-matched r tuples padded with nulls Database System Concepts - 5th Edition, Aug 27, 2005. 13.44 ©Silberschatz, Korth and Sudarshan Evaluation of Expressions  So far: we have seen algorithms for individual operations  Alternatives for evaluating an entire expression tree   Materialization: generate results of an expression whose inputs are relations or are already computed, materialize (store) it on disk. Repeat.  Pipelining: pass on tuples to parent operations even as an operation is being executed We study above alternatives in more detail Database System Concepts - 5th Edition, Aug 27, 2005. 13.45 ©Silberschatz, Korth and Sudarshan Materialization  Materialized evaluation: evaluate one operation at a time, starting at the lowest-level. Use intermediate results materialized into temporary relations to evaluate next-level operations.  E.g., in figure below, compute and store  balance2500(account ) then compute the store its join with customer, and finally compute the projections on customer-name. Database System Concepts - 5th Edition, Aug 27, 2005. 13.46 ©Silberschatz, Korth and Sudarshan Materialization (Cont.)  Materialized evaluation is always applicable  Cost of writing results to disk and reading them back can be quite high  Our cost formulas for operations ignore cost of writing results to disk, so   Overall cost = Sum of costs of individual operations + cost of writing intermediate results to disk Double buffering: use two output buffers for each operation, when one is full write it to disk while the other is getting filled  Allows overlap of disk writes with computation and reduces execution time Database System Concepts - 5th Edition, Aug 27, 2005. 13.47 ©Silberschatz, Korth and Sudarshan Pipelining  Pipelined evaluation : evaluate several operations simultaneously, passing the results of one operation on to the next.  E.g., in previous expression tree, don’t store result of  balance  (account 2500 ) instead, pass tuples directly to the join.. Similarly, don’t store result of join, pass tuples directly to projection.  Much cheaper than materialization: no need to store a temporary relation to disk.  Pipelining may not always be possible – e.g., sort, hash-join.  For pipelining to be effective, use evaluation algorithms that generate output tuples even as tuples are received for inputs to the operation.  Pipelines can be executed in two ways: demand driven and producer driven Database System Concepts - 5th Edition, Aug 27, 2005. 13.48 ©Silberschatz, Korth and Sudarshan Pipelining (Cont.)   In demand driven or lazy evaluation  system repeatedly requests next tuple from top level operation  Each operation requests next tuple from children operations as required, in order to output its next tuple  In between calls, operation has to maintain “state” so it knows what to return next In producer-driven or eager pipelining    Operators produce tuples eagerly and pass them up to their parents  Buffer maintained between operators, child puts tuples in buffer, parent removes tuples from buffer  if buffer is full, child waits till there is space in the buffer, and then generates more tuples System schedules operations that have space in output buffer and can process more input tuples Alternative name: pull and push models of pipelining Database System Concepts - 5th Edition, Aug 27, 2005. 13.49 ©Silberschatz, Korth and Sudarshan Pipelining (Cont.)  Implementation of demand-driven pipelining  Each operation is implemented as an iterator implementing the following operations  open() – E.g. file scan: initialize file scan » state: pointer to beginning of file – E.g.merge join: sort relations; »   state: pointers to beginning of sorted relations next() – E.g. for file scan: Output next tuple, and advance and store file pointer – E.g. for merge join: continue with merge from earlier state till next output tuple is found. Save pointers as iterator state. close() Database System Concepts - 5th Edition, Aug 27, 2005. 13.50 ©Silberschatz, Korth and Sudarshan Evaluation Algorithms for Pipelining   Some algorithms are not able to output results even as they get input tuples  E.g. merge join, or hash join  intermediate results written to disk and then read back Algorithm variants to generate (at least some) results on the fly, as input tuples are read in  E.g. hybrid hash join generates output tuples even as probe relation tuples in the in-memory partition (partition 0) are read in  Pipelined join technique: Hybrid hash join, modified to buffer partition 0 tuples of both relations in-memory, reading them as they become available, and output results of any matches between partition 0 tuples  When a new r0 tuple is found, match it with existing s0 tuples, output matches, and save it in r0  Symmetrically for s0 tuples Database System Concepts - 5th Edition, Aug 27, 2005. 13.51 ©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 Figure 13.2 Database System Concepts - 5th Edition, Aug 27, 2005. 13.53 ©Silberschatz, Korth and Sudarshan Complex Joins  Join involving three relations: loan  Strategy 1. Compute depositor (depositor customer)  Strategy 2. Computer loan customer.  Strategy 3. Perform the pair of joins at once. Build and index on loan for loan-number, and on customer for customer-name.  depositor customer customer; use result to compute loan depositor first, and then join the result with  For each tuple t in depositor, look up the corresponding tuples in customer and the corresponding tuples in loan.  Each tuple of deposit is examined exactly once. Strategy 3 combines two operations into one special-purpose operation that is more efficient than implementing two joins of two relations. Database System Concepts - 5th Edition, Aug 27, 2005. 13.54 ©Silberschatz, Korth and Sudarshan