Chapter 22: Distributed Databases

صفحه 1:
حووهرل() لس طلجت0) 00۱ بان 

صفحه 2:
+ Okt, 00: Onrburd Oxkiwen اه( مرا له سرا مت مها تسه س1 جاه( سس موه Ovarurreay Ondo it ‏طله) اه(‎ ‏ره‎ ‎Query Provessicnt‏ لها ‎Aeerxpoens Osrinted Dorie Oreviry ‏سره‎ ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 هوه‎ 

صفحه 3:
+ Ovrinted Ocibwe Opstew © dstibuted database systew ovusists oP loosely coupled sites thot shore ‏نموم مه‎ ‏امبر‎ ۲ ‏اوه(‎ systews thot ron vo cock site ore indepeudeat oP ‏اس‎ per © Prxswiew ‏و رو‎ dott of coe or wore sis Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 ووه‎ 

صفحه 4:
+ "cpa Dor ined Uidtwee 19/15 5 ‏سول لاد ستحصص حصا‎ © Ol sies have ideuicd spPiwore © Ore were of euck ober oad corer ty copperte tt procession user regents. © Cock ste surecders port oP ts cuivareny te teras oP right tt oboe schewas or svPuvare © Oppeus to wer us a sine systew © ‏ال لاس رما و‎ © OP Perec ‏وه له اوه ملهچ رون بط‎ ۱ ‏و( ما موی و ها اوه و سس(‎ query provessicry ١ ‏مس و مار و و صا جمسخامد دا سس(‎ provessicny بات جلخصمم ۱ و مسا و مور ‎Por‏ ای( لوا Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. coe ©Sbervehnts, Cork ced Cnakershe 

صفحه 5:
+ Ovrinted Daa Oorene © Qssuve ‏ما‎ data wodel © Repicaica © Gpstew wotutaics wuliple copier oF data, stored to diPPered stew, Por Poster rettevd ocd Poul ‏ماه‎ مه ۲ ‎Praxgoeuis stored in dstiuct sites‏ ۱ ای سا ی موه نی شش ۲ © Rektion ts portioned iti severd Prockreus: systew wotuties severcd ‏ام‎ ‎replicas oP rock suck ‏.سمس‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 هوه‎ 

صفحه 6:
@Octa Rephouva ۲ © rektivs or Proceed of a retitica is rephootied Pitts stored reduercacily to tua or اه و ‎stored of oll sites.‏ & متا ‎Pdbreptoaion oP a rektiog ts the vase where the‏ © ۲ Pulp ‏وال تلم‎ ore those te whick every sie opotaas a copy oP the ‏لول وی‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 هوه‎ 

صفحه 7:
+ Orta Repirion (Oou.) ۲ ‏مد‎ of Replication © Quokbiy: Pakire oP site cvotctotay relation edges ont result to veracity oP ‏جار‎ ‎replicas exist. © Cordktew: pers oo 1 wy be processed by severd wes ia porcitel, ۶ ‏را لورت و ماه توص مول لیلج‎ of cack ste ovotctoiog a replica oP r. ۲ Dirudvovtes of Rephration © deorewsed post oP updates! cock replica oP relation romust be upcaied. اس تسف و لس مسج :ای رو حدمت خأن بمج كوت ل ۰ ‎chia (kee spaniel ppuunreuny bound werk ore‏ مر با لوا روم جاتو امه تسج باب اجه روت رمع رو ی مات تاره بو ۱ ‎ra primary DOT‏ 1 Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مه‎ 

صفحه 8:
Oa Pragerciaios ای مه اس رن ريو رد عم مب ما ‎BE Over oF‏ ما اس و مور ‎Pronger‏ مه و و لس اه و ات تمس مرا ۳ ۲ ‏ای چا مه موسر لبون‎ ris spit fair severd sure ‏مرو‎ © ۵۱ ‏(رطموه ) روا سم موه و مه اس ماه‎ to eee ‏وم ز متس‎ © © specid ‏ع بو ات ال با رصاف‎ added to pads schews to serve ‏وا لو وه‎ مه مش( مج ما : ما ۲ ) اه اه مه سس ده ۲ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 هوه‎ 

صفحه 9:
+ ‏سس‎ Praqweddiva oP ‏مت‎ Rekiva bokrace SOO 999 99 ی bokrace 909 00000 ۹89 PSO ویو 909 999 CASS طمه_ مه Oar? 0۳9 0۳090 0999 سبحت اتوت جا Wisi Wisi ilside سسطهه سسطهه ( )یه سس گولا موه هوه 0 

صفحه 10:
مس( ربمت <ه بقل لبون تا | ص | جوم ‎q‏ مسا مارا 8 سور مارا ‎Odkeyrew ۳ 8‏ ‎Kaka 98‏ من ‎(obo 9‏ لسار 1 ‎Odkeyrew obs 89‏ لها سس رطا 0 ( طم وسجامجا بره مم سس ‎chet) = Tip‏ ‎bolo fyple_td‏ طمه_ موه 4 900 وم ‎e‏ 9909 02666 ‎oar? 909 9‏ 3 000۵ ۳ 9 99 0499 9 99 94۰9 90 9و9 ‎raf )‏ سار وه سا ‎Terns‏ = عمط ‎Orsdrer Gyre Orcewpe -O* Cre, gO, OOOO. ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ 

صفحه 11:
+ Odvadiags oP Praqeedidiva © Wortzword: © dhs pordlel provesstny 7a ‏مره تن وم‎ © dws a rektic t7 be splt so thot tuples ore located where they ore wet ‏تس مس سنا‎ ۲ ‏تین‎ ‎۱ ‎Prequediy wessed 8 pled uiibuie dows ePRipical ipictog of veriicd ‏حسمي ذا‎ ۶ ‏مر وم رس وم روط‎ ۲ ‏ای وا و و و له تن‎ |۱۰ be suoressivel Prepvedied ip oc wbirary dep. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 هه‎ 

صفحه 12:
رسیم مین + © Qos repay: Oewer to which ‏اه سنوی و رو ی مرو‎ he dletabs oP how aed where he dt tows ore stored faa detrbvted syste ۴ ‏ما و با روت ملس‎ t7! ۰ ‏رم ما‎ © ‏رم مشش‎ ‏رم را و‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 دوه‎ 

صفحه 13:
+ Darcey OP (Deis cow > Order 0. @uery deta tew cst hove a syetew-wide Unique ore. 9. should be possible to Pied the locatica oP data ews oP Picea. 9. Usha be possible to choo the bpcutiza ‏روص وا ول او‎ P. ack site shoud be oble to oredte cew data ews cuiconpusly. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 ووه‎ 

صفحه 14:
+ ‏اسف فقو‎ Gales = Dawe Ourter تست ۲ ‎of anes‏ وه و و ۶ ‎seamen: rere oped ch ne‏ لو لوم مد ولو ‎tee coh caer server ip bore‏ © تسده ۱ ها مس ری اس ۶ منت ۲ مت موی راو بت جحل © سا یوم بیس وم و ی اه ام وه سا ‎canoe Server‏ © Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 موه‎ 

صفحه 15:
Ove oP Okwes © Qlerwatve to ceuirdzed schewe! euch site prefixes ity owe site ‏روت ها تا‎ senve thot yeurrdes Le., site UP ‏وعد‎ © Cue hovieg 0 unique ideuiPier, cad avoids problews associated wit ‏امه ای‎ ۱ uetwork trresporeury. © Gohiion: Creuie a set oP dhkeses Por dat tews; Otore the wappirg of oles 1 the red ‏وی‎ of rack site. ۲ ۱ wer oe be ware of the physicdd lavaticg oP a dota teow, cro ‏ج‎ ‏لخاد‎ ۴ the dota tec ts woved Pro por site ty carter. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 دوه‎ 

صفحه 16:
جاح حدم !1 لحف اناا( + ۲ ‏و روت ی‎ dete of severd stew. © Goth sie hos 9 boo frocsuniiva women respousibte Por: © Dotetctetagy a boy Por recovery purposes: سس سا اه موه وه سا له و ‎Ponteipatay‏ © ‎of that site.‏ بجوم © uth ote hoe ‏اطع ات و مت و‎ Por! © Grontey he execuiiva oP trocsurives thot pricferaie of the site. ۱۳ exeruica. © Coordotiog the terwtodiog oP ead tracrariiva thot ‏ممعت‎ uf the sit, wwhick way ‏لآ اك لمصصاه عه دعاك أن اه ای با تمصع او‎ 5-3 0 e010 ©Sbervehnts, Cork ced Cnakershe 

صفحه 17:
transaction| manager Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. eon ©Sbervehnts, Cork ced Cnakershe 

صفحه 18:
+ Oystew Pure Ordo ۲ Pokres uae to dotted systews! © Coker of ante. © Lowe of ‏مس‎ ‎۱ ‏لوا‎ by cetvork trexewissira cocina protools suck a POPP © Boke ‏موم و‎ bk ١ ‏اموا‎ by wewwork probook, by novi ‏من تجو مسدب‎ cher cave kobe © Drtwork potion ٠١ 6 ‏وی‎ sail to be peticwd wheat har beea spl io buy or ore ca deg care fe beets ‏مجو وی وبا سوم و‎ ~ Onte: 0 subspsiew way ‏له له و اه یه‎ © Detwork portico ood sie Pokees ore yeurrdlly trdsticcquishable. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 دوه‎ 

صفحه 19:
Cow Prowovls © Cwm provok ore wed ty ere downy ures vies © co trocsuntiva which execules ot wulipte siies ‏هه ای اجه سره‎ sites, or uborted of oll the sites. © wt wreptible ty hove ‏د‎ roeeunivg cowed of vor site cod oboried ot ter Bl he ter peer court (PO) protec te widely ved © Ve tree-phose crow (QPO) prowl & wore eowplocted ond wore expeusive, but avoids socwe drawbacks oF ‏وذا "ا" اس ای وس‎ ‏ام ولج اس سا اسر‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. e016 ©Sbervehnts, Cork ced Cnakershe 

صفحه 20:
+ Two Pkwe Ovwwt Provo (OPO) ۳ Qoances Pobsioy wrdel — ‏وه راو وه لس‎ workin, cd do ot on sey her ban, suck oe epee mere sree pe obey sas ۲ ‏مش‎ of he protec f taticted by the coordicaior Per the ket step oF the trocsuntion kes bero reached. Bl Phe prowed Ruolver he bod ster of whi ‏اه مس‎ BE Let Phe a ‏ماسج‎ hited of ote G, ood fet ‏مه مت‎ ot G, ber 9 Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 21:
+ ‏سسجت‎ 0: Obracag a Deven ۳ ‏مس و ور و مه له من‎ © Cade he records <prepare > to the oa od Porces bog to skible stone © sek prepare TP wessnps 7 dl ster ot whick TP executed © Opera revevieg wes, iroccorivs woonger of site ‏او معط‎ the trocsuntion © Bet, ‏سره للم‎ > 1< to the by onnd seod ‏و‎ 1 weve ty O, ‏تا راو او و بخ‎ © add the record <ready T> Wo the boy © Pore dl revere Por To othe stone © sed ready TD wesene yO, Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 22:
+ ‏سا0‎ ©: be Oroeios ۱ Powbe cowed of O,receued a ready P ‏و بو وم‎ vies! ohenuee Tet be cbovted. BE Opordkntor ‏مت و طلی‎ revord, Spawn 7> or <cbort T>, Wo the bry od Pores ‏موه له و لو‎ Once he record stable sora to trevordble (eved F Pokies oom) © Coordentor seuds ‏موه و‎ ty eur portvipont tPprentay it oP the devisioa ‏عد د مسحي‎ © Ponteipaats toke approprisie utes broly. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 23:
+ ‏إن مدر[‎ Pahuee - Ote Pare Oka ote 6 recovers, texavices te bt detercoine he Pate of ‎Pofure.‏ هس سب سره مهو ‎ote exectes rede (7)‏ لس <1 موی ی را ۳ ‎۳ ‏لس <1 > مت را‎ ote executes wade (7) ‎۳ ‏ی را‎ ready 1> record: ste ant cout O, to ‏سب سم‎ of ۰ 1 ‏سس‎ reds (7) ۶ 1 1 ‏طم اس‎ )1( ‏تا لت ۵ ند و ‎records oowercikn‏ اس ‎ow‏ و ‎Bh he by‏ ما وم( مور و بطم 1« وه و سا ما۵ ماه( باس © ‎TP‏ مه اس سس ‏() طی سه سم 6 © ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 24:
+ Waaday oP Palures- Ovordertor Pale AP cordon Pais while ‏اور وی‎ Por Ps execute theo poricizaticry tes wet devide oa T's Pate: 1١ 118 ‏مرو ده‎ site ‏ویک و ی‎ 1< record in its log, thea TP oust be cowed, AP ‏عون اه صقصه مه‎ oc ‏بوک‎ 7> record te ts log, feo P cust be ‏واه‎ ۳ ‏برع ورن‎ portcipaicn ‏هوک همه بح جات‎ 1< record ta te bo, thew the Paled ‏وس 0 1 و او اه‎ ‏سل‎ T. (OB cece oP the chove moses holds, hea dl cote oes ‏اس‎ have ‏و‎ > 1۶ record fa her bxp, but oy ackdivad coil records (suck os <ebort > of <powwt T>). I his case wive ster net ut Por Op recover, to Pred ‏سوب‎ © @behtcg problew : ‏جوا چرشه‎ way have te walt Por ‏تم وا ی لطاوط‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 25:
+ Wardiay oP Pulures - Ostwork Portiva صم عدا اه سا روم سس و ‎poricipanis rewota‏ جا ‎of‏ لدت لدم صمت ‎Phe‏ ‎ef Pent oo the oot protocol.‏ Pike coordktor oad ts porinigoais betray io several partis: © Glee trot ore ol in the portiicg oootcioiey fhe coercion thick the coercion ros Paled, ced exerute the protocol ‏یی با ان وله جات لو وا‎ * Ov kare reste, but sites way stil hove to wait Por devisioa Prow coordi. © Phe ‏هت‎ od the ster ore th ‏ی اجه ماو‎ ١ the sites tothe vher portica have Paled, god Polow the usudl cocci protocol. ۱ ‏ری‎ ww harw results Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 26:
+ (Reccusry ad Oouneveccy Oot 19 eoubt trowurtows hue u Sready T>, but wether o <powet T>, wor oa Sebo T> boy record. و اه ‎oP‏ سوه وت بط امه همم با 0 ‎BL Revovery ckprikws pou uote bok Promos tothe lox.‏ ‎bet oF locks bell by T‏ = را ‎oP <ready D>, write ot Srewdy TD, L>‏ لس © ‎whe te bo writes (read broke mas be exited).‏ ‎ia hor‏ صاصم ‎Por every ede treceurtrs Phe locke‏ © ‎ready 7, L> bry record ore revexnired,‏ ۲ Rex ork rearquisiion, irrcsuniva processiay com res une; the coal or ‏لول و اسان‎ treacsurioas is perPorwed pvocurredly wits the exevuicd of Sew ‏اس‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 27:
4 Olercatve Oude’ oP Trawwiva Provessing BE Dotea of « sien ‏و اه ول وه مس‎ cerry ‏او‎ ولا ام و رون توا وق ۶ ۱ the to peri ot exterodky fottdted tracsucica to block food! trocsaniogs Por oc todeterotcate period © @lercaive wrdels cary vul ‏موی رطلوه نوا مج‎ © Code ty hoode wessuges west be vorePuly desiqaed to easure uiowicty ood durubliiy properties Por updates: ۱ ‏یی قاری‎ be cquarcaced, ta tat ioteroediate stages are visite, but ‏موی امه ون‎ on foozosisted ‏اه‎ resul due to ‏ووم مسجحجمت‎ امس عم نا موه و موه رس مس © مس ‎properties to‏ ی رسمه لبط با نا الم موی ‎Oessuyes ore‏ * ‎Kopleweutdtion techaiques kiter‏ مد ال ۱ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. e008? ©Sbervehnts, Cork ced Cnakershe 

صفحه 28:
Olercave Dudes (Ova.) © Ootvatoy excople: Pucds trowsPer between two books © Dun phos oot would hove the poteatdl to block updates oo the oor favolved fo Pues tracer تاه عمط © ‎ty oer ste‏ و و لو لو چم ‎worry Prow source‏ )+ اه فطل ‎oredis‏ لو مج سس )۱ © Deseagiay hee lrog bed used Por ‏جوخاوط مصيج) جصاصمصدم لاس‎ ‏هی سم‎ eri) سح رت ۲ © ‏لكيه ماص صصص‎ a Wessuye ‏سم وم راون ع‎ quorcaierd tr be delivered < Guorater os loony os desicaicg site is up ood reuckoble, code to bocce eadelverdble wessaqes wet dev be avaible ‎acco.‏ و با ما رو لو ری جد ججا امه أصدجه یو ,ول مها بطلمیه 1۳ © ‎ ‎4 ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 29:
4 Grror Contiow wik Perssirdt Desoagaq Bl Onde ty hands worsens her ty the core of vate of Palure sitios (eves ‏وی سم مه‎ delvery) ۱ sect bork to power site © Oke Palire wessuge i received Prow destoativa site, or destcaica sie tse bes unl exist, so7cey west be deposited back ia source ‏اجه‎ * Oroblew P source wormed hos bero dosed yet hues to take core of problew © ser code exevutog irocsuriog provessiry veto GPO does aot have to dead wi suk ‏عساه۳‎ © here we ony siuctiogs where ‏اه واه وت‎ error boa ‏ات‎ ‎beceht of obsrare ‏بسا اه‎ © Gx. prety ‏اس‎ ofl ‏اس‎ ures ‏اس‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 30:
+ Persea Dessay vod DorkPows Bl QorkPows prot o ‏لجيه اموسر‎ of inneunioad ‏له امه مس‎ sees eed pred ‏ی مومسم میرب‎ rete ogee © Gi, wheog book reveves o loon upplicdivg, way ceed to * Coctedt extercal credi-cherhiag ‏ره‎ ‎* Get approuds oP var or wore wes seed theo repped to ‏ما ما از‎ © Oe sin) worlPiws it Chopter OS © Cersisted wessugiay Porws the urdervicy foPrasticucture Por workAows ia ‏موی لاب د‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 90.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 31:
Cowurrewwy ‏اهبوهن)‎ موی لا و ی چا مره ای توص رن( ۲ Bl Qe woke thot ew sie poriopdes fo he execuioa of ‏اس او و‎ io eau ‏,اهاط‎ © De weve dl replicas of ony few ore updaied © ‏عجو لز(‎ how to rete this it cose oP site Puhires biter Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 32:
امومع موه رسب و و راوا رما ‎ste bok wane fot resides tao sie choses oie, sy‏ 0 موی سوم رقا ,6 ‎@ Okew ‏شم و‎ ceeds te tock a dota ites, de ct bak request te GS, rd locks swannger deterwices whether the lock coo be graced ‏رال‎ ‎۶ 1۴ ‏ما رس‎ weer seuds 0 wessuge ty the site whick taticted the request ‎© 1Pw, request ie dekved vail trea be graced, of which fee 0 wera bs ‏جا امج‎ the ‏له باه‎ ‎ ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ ‎ 

صفحه 33:
+ Orade-Look-Dacxpr Bpprowk (Ov) © Ve tenewive coo read the date itew Prow gay vor oP the sites of hick o replica oF the dott tes resides. Bl Ortes cnet be perPormed oa dl repos of otc tea BE @dhantoges of sche: ۶ ‏ماع‎ ‎۶ ‏سل م۵‎ borrdhrt 1 0 of ‏اه‎ ‎9 ‎© Ockercbly: syetew ‏مومت با تا ی ها‎ ote Pakire. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 34:
Ovrbuied book Oacaper Baths opprouk, ‏او ما جر ره‎ by lock wacuyers of each ite ‏جوم مورا‎ craic access ie ood data teow * ®t spenid protools way be used Por replicas ۲ ‏وت تمد‎ te destrbuted ced co be wade robust to ‏اه(‎ ‎۲ ‏مش امد نمی(‎ is wore cowphoued © beck ‏مش الم عون موم‎ + ‏مس سو()‎ this beter ۲ ‏اس هذا خأ من سوت‎ ‏موه و‎ © ‏ام رسب‎ © ‏اس لس‎ © Quon mee Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 35:
@riwary Ovpy Bl Choose oor replica of chat tow ty be the pricoory copy. © te ‏لحلاب جا مجارت حصا بصم‎ the pricary ote Por trot dot ‏مجو‎ ‎© OR Rereat dota tews con have dhRered privay ster ۲ ‏از‎ o eeuntod uerds ty lock o dota tee Q, trequests a lock of the priory site oP QL © Aeopteily vets locks co dll rephous oF the date ew © Oeeti © Coeur ovoid Por ‏لاو‎ data hooded sicviody to vorephouted dott - ieople ‏ماو‎ ‏ام( ا‎ اه اه ما مه امس و ۵ ‎1P the primary sie oP GQ Pus,‏ © .راوجمه با نویه لو و لین Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 36:
سسوسية 000 اله 19 ‏سای له ما ول لاك امه له مت ما مرا‎ recy Por ck tews stored of that ste. BE Okew o ixxeunton woken ty book oa voreploated this few GQ resid ot ote ©, 3 ‏مس او و و وم‎ © 12 © ‏لحطاصط جز‎ fn oo ee potble wode, theo he request tz delayed vel tt zor ‏اي سا‎ ©) Oke ‏وا سا ای سا مه مهو ما‎ warager seuds a wessuye back to the tector tochootoy trot the lock ‏اي وا اوح‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 37:
+ Ocerty Proton (vcs) dace of rephoded data ©4P Qe repioded oo tes, theo oooh request ‏و مه سا اه مج‎ swore trad oP oP the o sites in whick G ts stored. © Whe tensuciva does ont operde oo Quail thos obtaced ot locks ooo ‏اه سا لا اه رو‎ 6 © Ohea writoy the dete tiew, irresuntiva perPorws uries va oflrephrcr. B Beet © Cou be wed eves wheo sowe sites one ‏جا ادي‎ ١ details oo how ‏سم اج ارت وا‎ oP ste Phere hater B Orato © ‏تج‎ 6/6 + 0) woeseues Por handy bok requests, od (2 + () ‏وه سای شمه وا مس‎ ‎cack oP 9 trowsuriioces‏ وت - ما وه ات مه الم و( لو و ول و ‎sway have locks vo WOrd of the replicas oP‏ ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 جووو‎ 

صفحه 38:
Orved Proivvol boc lock ‏وچ اه ام و مت‎ wuiorty proieonl, however, requests Por shared locks ore korded dPPeredly frac requests Por exchusive locke. © Ohad bebe. Dheo 0 trocsartod ceeds te look dota lew Gt stop) requests a lock ‏مس‎ G Prow the lock wosager of vor sie ovotatotay replica of GQ. ۲ ‏مه مب( با مب‎ ceeds ty lock dota tee Q, ۱ ‏جن اه وج‎ Q Prow the lock ‏وی و أد اه وی‎ a replica oF Q. Bl @dhantaye - kopoeer kes overhead oa read opercioce. BE Derodvontne - ‏سمالي‎ sverheud oo writes Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 39:
Quon Oowews Prowvol ۲ ۵ ‏لت‎ of bok woeiy ocd biwsed promos Bl Guk ote & wokned u wekh © Let be the tod oP ol ote wei ۱ ‏من تال‎ ches ‏تحصو لحي‎ Q, ued wre gumnine ‏بي‎ ٠ ‏مات‎ Q+Q>G ud CTQ>6 ۶ ‏مس‎ vos be choses (aod ‏ی و اوه مس نا‎ tow Buk ‏و توت بابرا نس ام‎ thal he suc of he ote weighty & >= G, Bl Ck uric cnet lock punk repiow thal he suc of he site werkt ‏ري 0 << و‎ © Cor ww we wee dh replicas oe writes ‏ماه وه تاه وا تمس‎ iy be vouvaluble desorbed hater Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 40:
جیسب 4 لجسحاسصل جا لحب ححا جوت جامصصاصجم ام موی لوا م1 ۲ مد ی عوی و مق سا نس مه ‎Cock‏ © ماه لاس و و و و ‎yeuercie‏ صا وا تام و( © ۱ 0 unique Ipod ‏عدص مدص‎ pier o legicd couier or the todd clocks. © Clbd usique thvestap is obtaiued by courutecatay ‏جمصاددصا جوا سوم بط‎ wit the voique deciPier. local unique timestamp global unique identifier Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 41:
+ Droeokropiay (Ovc.) BO ote uth a eb cock wd wokm surder Roeser © Gil ded) correct sertdkzabhy ‏ماه مه‎ © Out “dowbestnes” rxewkow Bo Pix his probiew © DePe wahia pack ste 6, ‏حيصا د‎ cbc: (LO), whick qecercies the wine et eT © Rewuire frat © xvowe ts bird obok wherever ‏او تا سوه‎ عه ‎ber‏ يها بوي حا ء ‎weed‏ ی و ‎of LO,‏ سس © otis cose, site 6 advouzer tis Irgicdl clock te the vohue x + ۰ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. e000 ©Sbervehnts, Cork ced Cnakershe 

صفحه 42:
2 سن 0 + ۲ ‏تاه ما سوه لد لو توا‎ dua wk used dewees of ‏فلت رعا) تون‎ a sarosiee oP seri) Bu! cxeterckee replouioa! unkies are perPorwed ufo oknke “unter” ote, ‏راو و امسر له‎ tes. © Propogaivg is wt port oP the update ‏مجح تمس‎ * Ou be ‏ای شم اه للجم‎ * Ow be periods © Oa way oly be read ot shave sites, ot updated * Oo ceed te obtata lacks of cay rece site © )ةهقح ‏وموم كلها بمسحاسصل عونا لخامص بلط‎ ۱ €.4. Prow ceaird oP Pic to brook Pice © ‏رال صصص سوا أتخادى د85‎ quertes oP Ace Prov the rota dotdase Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 43:
+ Qeplowos wik Ded: Oowetewy (Ovu.) © Replies should see u ‏هوجو‎ suupshot of the database © Dhotis, 0 state oP the dotubase rePectog dl ePPevts of ol recsurtizas up i soe poi tothe sertdtzaiog order, ced a7 ePPevis of ‏عمجم تا رون‎ امه وه ‎@.y. Ore provides o Create Snapshot stewed to rede‏ ا ‎site‏ و و او ما ‎oro set oP‏ ما و ان 0 ache ۶ ‏یه‎ rePresk (cooks or periods) or axnnnd rebresk Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 44:
+ 6( ‏ای ی بط‎ bans Rapin Bl Qik wulkraster reptoaion (doo caked urdie-ce here rephrioa) updos ae ‏اه و ار اتمه وه لو مور روف تم‎ © ‏سنوی بت مس ان بل لا و اون مه‎ of the ‏تب‎ oP rephoatiog, ced chtubose ‏مومس سل سوم مسر‎ of ‏اس سیر بت با‎ > Coupled uth 6 ‏موه سم‎ © Oey sees suppor hey propio ‏اجه ی بات‎ Per genset ieee © @lows updiies to voour eve P sows sites ore dscvouenied Pro the ‏,انم‎ bul of the cost oP ‏ری‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 موم‎ 

صفحه 45:
ما سین + | :© جد د ‎Dy‏ مس امن ۷ ‎cen tow‏ 1: 00 ware () rie (”) ware (X) | | | ‏ماس‎ ‏اسر (00) سس‎ ‏جسن‎ )0( rat ‏ام را و‎ Wott Por X4ork oa Y (Resch ‏ام و موی ای اتقو‎ fools ot ether or ‏سپس‎ i Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 46:
Ovurdized Opprowk ره وه و و ای له له و اي ‎qebdl watPor‏ © ‎coor dictation‏ موم © Red gopk: Red, but wckorw, state oF the sysiew. ۱ the ovaiciler charter the ‏اجه اه لحم‎ ات لصو سا ی وی خن ای با ۳ عوك مت اما جهن مت لحم و لو ‎is‏ لو رو هن عون ۱ .وماعواصل-صاعين عطاصدا ها جلججم مادصل صمت ددا © 18 ۱ ‏رو وت‎ ft selects ‏اه لت له مش و‎ sites. Nhe sites voll back the victor trorcsuntion. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 47:
site Sy Global site Sy Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏موم‎ 

صفحه 48:
Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 49:
+ doe Oper (Ova) Bl Guppoee tht starter Pro the ote show it Pure, ٩ 1 ‏مه مس‎ ot G, > renin Wo were eee TT ‏مسج عم مج‎ Dornnpr of ste G, to the coordkntor) ©. Bud hea 1, requests o resource held by Tt ote Ge ١ ‏مه و ماج‎ Keer TP Brow Ga to be coordi ۱ ‏موه بو سا ما ید سین‎ reaches bePore he debte wesecne ۶ ‏وال ان و صط جوا من با‎ ۲ ‏اوه هط لا با‎ ‏وه مر م2‎ BL Dhe Babe oye cbove ‏و ای جر‎ rey. ۲ Rabe cides ‏الج جا اما او مج مس‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 50:
+ ‏پوس(‎ Rolohe © Qoevessey rolbacks way result wheo deadock her todeed ‏و وه تحص‎ ‏مش‎ bos ber picked, ced cecouwhie gor of the troczantizes we oborted Por ‏وی‎ varehted to the deudioct. © Ocevessey rolbachs coo result Prow Pobe oyles te the qobdl watPor graph; owever, likelihood oP Pabse ‏ناج طسو‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 60.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 51:
بلس igh ‏وه واه‎ Por whick systew is wt Puy usable should be extrecehy baw (ext. 99.99% carb) موی ‎oP‏ و( خن اد مشخ نا مرو ان روا ما( © (Polures are wore likely to horge distributed systews ۲ ‏سوواط و رات سا و‎ wet مه سی) ۰ سوه رو موه وه مرو با ملس ۶ اه ‎or bik te‏ وه ارجام © BE ‏سا‎ detevioa: dofoquichioy koh Poker Brow sie Paki ix hard © (port) sohtion: ‏مادا جد © باصملا جا دامتعا عون سره‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 52:
+ ReooP kurta طسب ۲ اه تاه هه وت ما تسوا اه ‎bor‏ © بوم روا سوه ام ان لت موه لاله اسب مج بجبله() + اه اه من ‎old lok‏ ‎incase ody soe repioos oP a dott itew Poted, tay be‏ نصا ‎hed accessed data of o Poled site‏ ما شم سامت ‎possible to‏ ‎(ces ate)‏ ‎© 1Prephcoted dota tews were of Poled site, update sysiew mukioy to rewove thew Brow the bet oF replicas. ‎* Dhis should be reversed wheo Poled ste recovers, but addiiccal care weeds ty be tobe ty bray voles up to ‏ول‎ ‎© (Po Poled ste wos u pedir server Por sowe subspsiew, oo period wust be held to deterctor the cew server ‏مس الما جاص ,فل ممت بروج تتجهمه ‎C.y. cere server,‏ * ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 53:
QevodPPiquraiod (Ovdi.) © Gree wetwork portion way ont be detiequichable Prow site Poakire, the Polowtcry ‏مد‎ oust be avoided © Ww oe wre vedi servers elevied in distort portico © Oore hoo coe portion updaies a reploded data few © Opdaes net be ‏و اه موه مره ی معط‎ dou ۲ ‏مسرت‎ worry besed ‏ام‎ ‎۱ coe write ol wvalible” ip trotdkziay but causes probes Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 54:
+ Dopriy-Besed Bpprowk ۲ ۱ wapriy prowool Por detibuied poaruneuy ordi co be wordPied to work even P soe sities ‏جاح صمي جمد‎ © Gack replica oP euch tew hos 0 version auewber which is updated ‏ارت‎ ‎replica if updated, os vulloed bets © bch request & seu ty of ews 4 the ster of which tew rephour ure stored ‏ددمت اعون ل‎ Wy wheat a back is ob [ ‏اه رزوی و وی‎ © Read operatives bok of oll rephoas locked, ood read the volue Proc the replica uth: largest versiva ‏او‎ © Dav urte this ude oad versiog eanober back ‏اس مرت سا‎ ‏اه مس‎ (ao ceed tp obra locke oa ol repos Por thts task) Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 55:
+ Dopriy-Besed Bpprowk ۲ Dopriy proton (Ovu.) © Orie opercticns * Brod highest versiva ounober like reads, cod set opus versiva ‏لام وا اوه‎ highest versica + * Ortes ore thea perPorwed 7a ll locked repioos ‏من امس و اج‎ these rephous ‏تنج طا اعد جز‎ versica ‏او‎ © Cakwes (sework oad ste) cose ‏جه مما جه ام مه‎ ۲ 3 ‏اه و اه ری و ی وی و‎ ey updated cute ites > Duras reads 3 wopriy oP reploas ore ovakible ‏اوه ی لب و‎ < ‏رس ما م۵‎ 8 phwe aww oa be wed io ‏ترس سل‎ © Ove: reas ore quarnived ip see kiest versie oP cot tea © Reteyraion if trvid okies ceeds ty be coor © Quorn oer cherihew vou be skvtely ‏الم‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 56:
+ Read Our Orie @l (Buckle) © Oreed prowonl fo sperid cose oP que poe © Ohaus reads ty new! oxy ove replica bul ukies require ol eplows ty be ‏جاج امي‎ ot oon eve (caked read vor write ol) BE Read exe write dl avakdbe (iyporkiy Puled otes) t» uliruive, bu ‏محص‎ 04 Poled ak way come buck up, witout a dervouerted ste ever beter awwore tof twas dscoccenied © Whe site hea hos oh volves, oad a read Pow thot site woud returc ot ‏سام موم‎ ر ام ما ما ال مر اه( اه رنه رن طد 1۴ ۶ سوم مس © Oth cetwork pontiionicg, sites to cock ‏روت ماو‎ update sae feo 02000000 © bebevieg sities to oer portiiccs have of Potted Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 57:
دمتسي حا ‎Ow‏ + 1# Qhew Poled ste recovers, tet cock up wits oll upddtes thot it wiesed whe t wes dow © Probe: updates way be happecien ty ewe whose rephod te stored of the site uhte the site is recoveries © Goktion (: hot dhupdites 7a systed while ‏د وس‎ sit * Ooncceptible derupiog © Gokitos ©: ‏او للد عامط‎ oP ol dota tecws of the site, update to hitest versivs, thea retease locks * Ober scluticas uit: beter pramunreany dso ‏جات عليه‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 000? ©Sbervehnts, Cork ced Cnakershe 

صفحه 58:
جیلعهظ) روما زر موه موون) ص اج ‎syste (Gevira (PD) are obo‏ (سوه ابا) جاص س0 1 ره ۷ ‎provide‏ اوه سا ما له او بو و وی ی ‎Bl‏ ‏۱ ‏او سر 6 را رصن لس ‎Wo weed Por‏ © ها موس وی مس نو خن او نت ‎BE Doky deirbued chicbwees‏ ام رویط رت اجه سل )© >( ‎rules‏ ند روا تاره ‎seemed veh reser‏ بت وولو و اجب میب بو اجره با ‎backup syste‏ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 59:
+ Qoordertor Oubotra مه 5 ۱ tPorwaivg lool ty ossunve the role oP ‏لو‎ P the ‏یی مه‎ اه ام سوه سا مت له ول سوه ها ‎exentes‏ © جما جد ‎state toPorovatiza‏ موی و( لت امه ‎fPorwatiza os the‏ ‎atu coercion‏ © dws Post revovery Prow ooordtor Pohure but evolves overhead curt sored provesskny. ۲ ‏سود منیا‎ © eed tp ‏هه و بنج و اد‎ oP Pokies: © Cx: Buly Okorther - applicable to spstews where every site oot ‏لمجو‎ a weep ty every ver sie. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 90.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 60:
لسب) راد ۱ ۳ ۵ ‏همم‎ request thot te oot cxewered by the ooordkntor ‏اه اس‎ fiend 7, weer th the coordkrtor kar Paled (G, tes to rect too ce the ew ‏ور‎ Bl Ose aks on ‏اه ما لاه نت جنات وه تا وی ماه‎ ۵ thea wats Por oy of hese provesses ty oxewer wie ۰ BR resper wthia 7, connor hot ‏لك‎ ster wih ‏لامجاي‎ yreder thos fave Paled, (,elerts teeP the cw cordon. BO Rewer 6 reveurd ©, bec thor ‏صمب , 117 مسمجنا‎ io receive a ene heer ie a bebe ‏سجس لو سوام‎ erie bers br ‏سییر‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 61:
®ub Okprikee (Ovu.) BP co wessuge ts seul wikia 7", oss the site wily o higher oeober hos Baked; (©, restarts he ‏سحي‎ BO Per o Poked ste recovers, ١ ‏اجه مش سا امه‎ sure ‏اه‎ ۲ ۳ ‏راجت بو مرو مه مه ی‎ he revovered ste Porves oll processes ‏با با ای نوا نت‎ t becowe the coordadior site, eved P there ‏ای اه نت وه عشه رام و‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 62:
+ Due third Gaaey (Proceso © Por cedized systews, the priwery orterica Por ‏واه اه سا مس‎ ‏رود لاور‎ ts the cucber of dish ‏یت‎ ۲ oodstibuied spstew, viher issues cust be tobe isto oat: © Dhe opst oP a data trowswissiva over the ueiwork. © De poteoial yeia tt perPorcaure Brow kevioy severd sites provess parts oP the query ta parcel. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 63:
Query PreawPorweips BP renshitay dkebroic queries vu Prarenis. © 1 ‏ماو موی وا عاتاوومم سا مه‎ r Proc its Procevenis سم ما ماه مه ‎Rephive relative rby the expression to‏ © ‎fate‏ ماو میت سا اه موه اما سا ین ۲ ‎yack cane abn (OPEV )‏ 0 > سحت ‏(سسص) ‎ ‎OO =O ‏مرن سپس‎ DRE ‏ال‎ ) bere 6 ‏موی‎ = “ds” (vere, U arent) hick ts opterized toi ‎© pene ‏لا )مریم یس‎ ٩ ‏سیر‎ ares ent (PAM) ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 64:
+ Gece Gary (Oca) © Gree upon, hos oly tuples pertotcten to the ‏واه ی هب مها ابا‎ the ‏وه مساو‎ ماه و رات و ما سا رانا ‎aye, (et)‏ دا ‎BOD ip expressive tp the exopty set recardiess of the orutects oP the oppruat rebates,‏ ‎urepunl, os the result oP the query.‏ وا ‎B® Pied strate & Por the Wilde site to‏ ‎Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 65:
Greople Join Provessiay ۲ ‏لین‎ the Polo wien rebaicod debra expression to whick the three rekiiccs ore ora gery bred o ote G,, he syotew ceeds to ‏سم‎ he rem ot ote 6۵ سیم با نمی سییر یی ‎br tet:‏ سسط مه ‎stored ot ste G,‏ & تسه 8 سس 6 ملسم Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 66:
+ Bosebl Gravy ۵ ‏رصن‎ 2 Bhp copes of ol bree rehtio tp ote G, oad choose ‏و‎ irdew Por proveseknt he ectre broly ot st G; Bhp oon of he corn retiod i ote 6, ond cowpule fore, = arnt Go. Chip ear, Prine Gy Ga, ood orp ter, term, broek Oo. Chip be rend trom, 7 Oy. Bl Dever sicker sircieyer, exchuayiay he ew G,, Go, G, Bl Dest oowwider Policy Portors: اه رال و مس ۶ © pst of ieee a dota block between ‏اه‎ © rebive provesskn speed ut ack ite Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 67:
obra he remit ot GO). x re Coewipts Oruew BE Let n be orcktioa wik schewa (R, stores ot vie G, vires vie G., Let ra be arckioa wik sckewa R., Bl Cucknte he expressira ry, Pend 1. ‏هو‎ term — Mao ae (rl) ۷ ۰ 13 ©. Chp uy, row Gr G. 6 Crm beg rg carla, ۰ B®. Ohy tur, yo ©, b Oy, a 8. Cowie eer ot G). Dh he soe ‏ص‎ e007 Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 68:
Cored OePraica Bl Phe sewipta of ry wih r,t dewrted byt re bx, لط لسمجاهجت فصا لخن سارو ما ارم Toe this deProed by: Taw (ra rs) عم ّ خام ‎co be exteaded to 0 series‏ ره وله با ,لول هروه ۲اه عم ‎(Por‏ sewipia steps. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 69:
+ dota Orreterios thet Bipot Pardo ۲ ‏سین‎ ey Mg, Vn, rg here retain rt stored ot ste GO, Phe resut ‏اصح‎ ‎be preseuied ot ste ‏بر‎ ge ede ‏اعد يك‎ re reed a Oe reer ‏ع اسيك عا وه‎ Cet ee aC, ۳ BG, chips hikes oP (ry ra) Oye hey proche Be 5 ‏اليك‎ re) Gy ۸ ‏لد‎ و ام ام ی یر اه معط لاب امن وین (e wae. Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 70:
ی () 2222222522 + © Oc dotubasr upphodiows require dota Prow ‏سل رس و روم و‎ located ‏شا مها و‎ of hardware ood soPivare phiPorns (عت , ان راما بل رورت لت ما (Decrees secret peers ‏سب‎ be ‏سپ‎ Oper rewy oro way be besed ‏اس مج سل مه‎ ‏لصف مومس‎ BE Gyseurbvel detals chovst perio oe toy Keowpuible. ۴ ۵ ‏موه هی‎ & u oeP vere kiver ow yy oP expiry ‏برد مق‎ ‏اسان‎ te descend to nmap ite ‏ناج ركام‎ © Credes on their of bud dakbuse ieyruiza wit ay phyoid database tect Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.70 ©Sbervehnts, Cork ced Cnakershe 

صفحه 71:
۱ ‏موی مورا‎ od ukrivistaive ovoid lows use ‏اه‎ spevic-purpose DOOGs ۲ Gey wank 0 iPied bros DOG © CUlneqraten iy u owns DOG Rare » Deckard ‏لمي بجو مايا‎ opt of ‏سوم صر‎ > Orgnizancrdlpolicad WP idies ~ Orprizatons de oot wort io Wve up cecil oa tet det — Dood ‏یل‎ wk ‏جا‎ rete a qed ded of ‏موه‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 72:
+ OvPed Ow oP Ona BQ creewedt oa a vr ts cored © Dredly the rebated coded Bl Oqeewedt oa a ora creed shew © PReredt cower Por exxoe ‏اه لا‎ ‏بو سوت موم اور با‎ Bl ‏و و وه‎ represectatioa oP shared det ۶ ‏وا مد بر‎ prevbivn, ۶ ‏وه سول‎ ١ ۵8۵0۲ ‏مس‎ 0 > Gort order varios Bl Ocreewedt oa wits oP eure © Octo to anes © Cy. 2 ve Ovkxpe, Ouobt ve Booby Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏وجوه‎ 

صفحه 73:
Query Provessiag ۴ Geverd eves query processing tau heterryeues doth © Gchews ‏ماص‎ ‎© Orie a wrapper Por pack dott source to trocekte dota to ‏ممجماءد اجام‎ © Oreppers cost obey trocekte urcktes ‏من‎ debdl sckewa to undies ot bool sokeur BE Lioted query vapcbiies © Gowe deta sources dw ody revircted Pores oP sevice ١ Ban. web Pores, Plat Ble dota sources © Queres kee ty be brokew up ond processed pody of hee source ocd party ott Ha cee ‏سم موه ماه وان موه موم با پر‎ © Decide whick ster ‏بو سم و‎ ل لا سا0 لح 0 لا سواه 1 وجوه 0 

صفحه 74:
مس ین + وه مد س مسا طوطی سب نكا ‎systews‏ وه ود 0 7 by proviker ‏قاط اتب مه‎ view, od providers suery Pores x1 dobal view ۶ ‏سل(‎ Bul Recked crutdatcbase syste, wedhiors qecerdy dy ot bebe ‏اعورم جو مسجو عمو‎ © Ouke teres weddior ued wulidctobase ore suweives used toter ‏سس‎ © Dhe tere virtud dotdbere ty deo werd to rePer ‏هلسلس وا‎ ‏مدرد‎ Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 هوه‎ 

صفحه 75:
Orsviry Opsews © Dyed hieds of ‏مت رو‎ © Cropbpee ‏اس كلما‎ os ove, id, ewal, phour, oPPive addr, .. © ve persvod ‏ام ول مس( توت ها تا موه‎ ٠ ‏.دج‎ browser bovkwarks © Ohte pes © Cortes ‏ی بر سم‎ or tbat ۱ ect Por Porword lookup io Ped wore ‏و موه‎ © Velow pages © Cottes organized by properties © Cor reverse bovhup tp Pied cuits wotchioy sper Pic requirececis © ‏موه لت چا نا وه سس میا(‎ on ‏ات‎ ‎© Bhernive 0: Deb isterPace. Dot great Por progres © Clernive ©: Gperidzed dreviory were protools * Coupled wih spevtdized user tterPaces Ez Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 وجوه‎ 

صفحه 76:
+ reir) (Docsee (Precnol Bost waxed) wed drectiory ‏تسم چیه‎ © LOBE (ikwuerh Drevin ‏اس وه‎ ۶ Groped Prow eater (SOO proc! ۲ ‏میت‎ Ohy wt wer chtcbase protools the ODBOMDBO? 9 ‏تسه‎ © GropiPied proces Por ‏لاه‎ ype oP data urcess, evolved parcel to 000000 © Crovite ‏وه و‎ Kerarchicd cecviey wechovisw sili ty Pie systew ‏و‎ <١ cua be porticaed ceooupt wutple servers Por dPPerect party of the Herarchy, yet vive oi sive view ty eer - oy. dPPeredt servers Por Bel babs Dura Wil ood Bell bobs unpre © Orevioies woy use dotcboses us storage werhoisw Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 وجوه‎ 

صفحه 77:
4 LO@®@: bigkweidt Oreviry Bovess Prowonl ۲ ۱۸/۵۵ Ou Dede © Oct Occupation © Osrinied Dreviory Trees Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 007 ©Sbervehnts, Cork ced Cnakershe 

صفحه 78:
LOOP Ode Dodsl © LOOP ‏اه سس سس‎ © Cores oe sober ‏اه‎ ‎۳ Gok ery cnet hae wine ‏چم لام‎ )00( 0 cnt wp oF a sequeue of rehive detached canes (ROOz) BE Gy. Fo OD © one Ciberechaz, met Lobe, valcect, POCO © Granted ROOs (oom be speoPied we port of svkewa) ( ‏مرو زين‎ RRO OU ‏قن سمي‎ 7 مس نو ( © Grokor to paks too Pile spstew but writed ta reverse directa Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 79:
+ LOOP Oca Dodel (Ovw) B Gores ma hae atrbtes: © Otrbntes ore ‏اس رجا لجس مساج‎ © LOOP kee severd butts rer * @rcory, sire, toe types 2 Dek ‏لماص مااع‎ PoptikDckhese! postal ockhene B LODE chws dePrition oP vbint chose © Obert theses speck y ‏چا له وی جات‎ © Con vee iohertowwe ‏ماه ماه سای نا‎ © Cory von be ‏لاه‎ ty be oP var or wore oben choses > Oo weedy hue safe wostepertic ie Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 وجوه‎ 

صفحه 80:
+ LOOP Orta Dorel (vvw.) © Gottes oryavized into 0 dreviory tPoreuiod tree uocordery to ther DOs © beh tevel wudly represeu speviic vbievts ای متس ‎objevis suck us‏ موس وی ‎Aotercal ode‏ 8 هس نو © ‏اه مان‎ a onde tohert the OO of ‏لت رتسم‎ ocd x RODS ۱ €.q, taternd ode wit OD 08 ~ ‏یه ان‎ have OO starter wits D=OGO aed Rutker ROOs suck us ‏و‎ pr ou > OO oF oa eciry ex be yevercted by traversisny pk Proxy vont © beck evel can be oa dks pottery to coher ect > Cottes can tus hae wore fro oa DO ~ 6.9). person wore hot poe ‏ام‎ it Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 81:
+ LOOP Orta Dactpukiva © Ocdke GOL, LO@® des owt dePice DOL or DOL. 198 estewd, it dePices 0 vetwork protoonl Por ODL ood DOL © Overs wer 0a BP or ‏حلمب مت متس مس‎ © LOO ‏و( ان سا سل‎ < ۱0۵ Dcta Ieterchorcge Porat (LOW) ۲ ‏موی رصن‎ & very stop! voy selevion & projection Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 0.00 

صفحه 82:
+ LOOP Crores © LOGEC gery wet seo) © Buse: a ade to the OVP Prow where search is te stot ۶ © searck ‏موه‎ ‎* @otecs cow biraicg oF ‏اه مس لس‎ oP ‏ی‎ ‎~ Cxuntiy, wikhoords ood opproxtcoie equally supported © © ‏عد‎ < hist the bose, he bose oad ‏عت ,مط احات كا‎ he cutie subtree Pro the base ای سا نا هه © 1۱ va cucober oP results ced oa respurce DoCERDpICT © Dey ob spePy whether ty aiewotcdly derePereure uhases B LOOP ORbs we ove uy oF spevihyiog query © LOOP OP is cawher chercaive Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 83:
+ LOGE 06 Bret ptf ORD sper server und DO of bose © hop: /rareseorch.belabs crmlp=buced,c= DOD ۲ Optrad Parker pars separcted by 9 ‏اسامد‎ ‎۱ Pra Kort © Optoud pats spechy 6 atirbutes to returs (scopy crea ol) Goope (pub ‏وله‎ putes subir) 2 Gearck ooaion (ou (ork) Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 84:
+ O Onde wig LOOP OP1 #include <stdio.h> #include <Idap.h> main() { LDAP *Id; LDAPMessage *res, *entry; char *dn, *attr, *attrList [ ] = {“telephoneNumber”, NULL}; BerElement *ptr; int vals, i; // Open a connection to server Id = Idap_open(“aura.research.bell-labs.com”, LDAP_PORT); Idap_simple_bind(Id, “avi”, “avi-passwd”); ... actual query (next slide) ... Idap_unbind(Id); Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 85:
+ O Onde wry LOOP P11 (Ov) Idap_search_s(Id, “o=Lucent, c=USA”, LDAP_SCOPE_SUBTREE, “cn=Korth”, attrList, /* attrsonly*/ 0, &res); /*attrsonly = 1 => return only schema not actual results*/ printf(“found%d entries”, Idap_count_entries(Id, res)); for (entry=Idap_first_entry(Id, res); entry != NULL; entry=Idap_next_entry(id, entry)) { dn = Idap_get_dn(Id, entry); printf(“dn: %s”, dn); /* dn: DN of matching entry */ Idap_memfree(dn); for(attr = Idap_first_attribute(Id, entry, &ptr); attr != NULL; attr = Idap_next_attribute(Id, entry, ptr)) { // for each attribute printf(“%s:", attr); // print name of attribute vals = Idap_¢ ‘get. values(Id, entry, attr); for (i = 0; vals[i] != NULL; i ++) printf(“%s”, vals[i]); // since attrs can be multivaluec Idap_value_free(vals); 1 Idap_msgfree(res); Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏مومه‎ 

صفحه 86:
+ LOOP OP1 (Ovu.) © LOOP OC 1 dev kes Pucrtiow to cree, update cod delete ‏تفه‎ ‎§ Gock Pusntiva col behaves os o sepordie ‏و‎ ‎© LOOE dees wt support aiowiciy oP updates Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 00.00 ©Sbervehnts, Cork ced Cnakershe 

صفحه 87:
+ Ovrinied Orevory Trove © Orprazatzad ‏روت مق وت‎ be spit ict ‏وصصما موه تس عمط‎ ۶ ‏له‎ DP a OID ques ROO to be tagged a to ol euites to get oa vera DD » Cy. wo O's, vue wikk subi p=burect, 0 ‏ماک ,اکن اه کت اس لت‎ ۱ ‏من لوا 0۲۵ مب امد اه تم‎ yevyraphicd locatica or by oresrtzoizcrat ‏تسه‎ ۱ replication (woster-skve or wulierster rephoctio) oF OIDs (cet part oP LOOP © stecrkard) © Bak ‏ات وت و وا ورن با رون )و و‎ OID © Cy. OF Cet bobs way have a separde OID, oad OID Por p=bured way hove a ‏سا‎ wa n= @ell Labs coutaictay a rePerrd ty the Bel bobs DV © RePerdbs or the hey iy toteqrates ‏تست و ماه تال و‎ © Okeo a server yes a gery reuchieg a rePenrd onde, tay ether > Ponward query to rePerred OP eed retire ‏رجات صا جحت‎ oF * Give rePerrd back te cheat, whick trowsporeuly seuds query to rePerred DIT (Wako weer ter etic) Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 88:
Gad oP Okaper 

صفحه 89:
نعمرسمحوه ی ام ۳ ۶ © tony por, ot eet owe ste wast be up. © 0101 ‏دده‎ (C vies (patiepads oe well w coordkrapr) om Pal Phove (: Obracioy Prekoinay Devtivs: ‘Meindl ‏و‎ 000 ۵ ۰ © Cuen step ren ty cowl P eeiruied ‏ود ص طا‎ Prove © oF OPC f oplt tay © phases, Pose © oad Phare 9 oF OPC ۰ eed revords tia wuipk: (ct lewst (K) stew © 41a prose 9, coordknior seus ‏ابوج اموت‎ wessace i ol parioivatay ster, Qader OPO, honed of pre-cowsi deveiva cas be used to cowent despie coor diraior ‏اه‎ © unk blockioy problew os bag a7 < (C sites Pal Orautraks! © higher overheads: © events way ool be satePed ta practice 4 Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 90:
Prue 69 مموو Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. 

صفحه 91:
0 سب 3 Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 92:
coordinator Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. ‏سا0 لح 0 لا سواه 1 مومه‎ 

صفحه 93:
Oxsdrer Gyetrw Oneewpe -O* Crim, Gui O, OOOO. Prue ‏م66‎ 

Chapter 22: Distributed Databases Database System Concepts, 5th Ed. ©Silberschatz, Korth and Sudarshan See www.db-book.com for conditions on re-use Chapter 22: Distributed Databases  Heterogeneous and Homogeneous Databases Distributed Data Storage  Distributed Transactions  Commit Protocols  Concurrency Control in Distributed Databases   Availability  Distributed Query Processing  Heterogeneous Distributed Databases  Directory Systems Database System Concepts - 5th Edition, Aug 22, 2005. 22.2 ©Silberschatz, Korth and Sudarshan Distributed Database System  A distributed database system consists of loosely coupled sites that share no physical component  Database systems that run on each site are independent of each other  Transactions may access data at one or more sites Database System Concepts - 5th Edition, Aug 22, 2005. 22.3 ©Silberschatz, Korth and Sudarshan Homogeneous Distributed Databases   In a homogeneous distributed database  All sites have identical software  Are aware of each other and agree to cooperate in processing user requests.  Each site surrenders part of its autonomy in terms of right to change schemas or software  Appears to user as a single system In a heterogeneous distributed database   Different sites may use different schemas and software  Difference in schema is a major problem for query processing  Difference in software is a major problem for transaction processing Sites may not be aware of each other and may provide only limited facilities for cooperation in transaction processing Database System Concepts - 5th Edition, Aug 22, 2005. 22.4 ©Silberschatz, Korth and Sudarshan Distributed Data Storage  Assume relational data model  Replication   Fragmentation   System maintains multiple copies of data, stored in different sites, for faster retrieval and fault tolerance. Relation is partitioned into several fragments stored in distinct sites Replication and fragmentation can be combined  Relation is partitioned into several fragments: system maintains several identical replicas of each such fragment. Database System Concepts - 5th Edition, Aug 22, 2005. 22.5 ©Silberschatz, Korth and Sudarshan Data Replication  A relation or fragment of a relation is replicated if it is stored redundantly in two or more sites.  Full replication of a relation is the case where the relation is stored at all sites.  Fully redundant databases are those in which every site contains a copy of the entire database. Database System Concepts - 5th Edition, Aug 22, 2005. 22.6 ©Silberschatz, Korth and Sudarshan Data Replication (Cont.)  Advantages of Replication  Availability: failure of site containing relation r does not result in unavailability of r is replicas exist.  Parallelism: queries on r may be processed by several nodes in parallel. Reduced data transfer: relation r is available locally at each site containing a replica of r.  Disadvantages of Replication  Increased cost of updates: each replica of relation r must be updated.   Increased complexity of concurrency control: concurrent updates to distinct replicas may lead to inconsistent data unless special concurrency control mechanisms are implemented.  One solution: choose one copy as primary copy and apply concurrency control operations on primary copy Database System Concepts - 5th Edition, Aug 22, 2005. 22.7 ©Silberschatz, Korth and Sudarshan Data Fragmentation  Division of relation r into fragments r1, r2, …, rn which contain sufficient information to reconstruct relation r.  Horizontal fragmentation: each tuple of r is assigned to one or more fragments  Vertical fragmentation: the schema for relation r is split into several smaller schemas  All schemas must contain a common candidate key (or superkey) to ensure lossless join property.  A special attribute, the tuple-id attribute may be added to each schema to serve as a candidate key.  Example : relation account with following schema  Account = (branch_name, account_number, balance ) Database System Concepts - 5th Edition, Aug 22, 2005. 22.8 ©Silberschatz, Korth and Sudarshan Horizontal Fragmentation of account Relation branch_name Hillside Hillside Hillside account_number balance A-305 A-226 A-155 500 336 62 account1 = branch_name=“Hillside” (account ) branch_name Valleyview Valleyview Valleyview Valleyview account_number A-177 A-402 A-408 A-639 balance 205 10000 1123 750 account2 = branch_name=“Valleyview” (account ) Database System Concepts - 5th Edition, Aug 22, 2005. 22.9 ©Silberschatz, Korth and Sudarshan Vertical Fragmentation of employee_info Relation branch_name customer_name tuple_id Lowman Hillside Camp Hillside Camp Valleyview Kahn Valleyview Kahn Hillside Kahn Valleyview Green Valleyview deposit1 = branch_name, customer_name, tuple_id (employee_info ) account_number balance 500 A-305 336 A-226 205 A-177 10000 A-402 62 A-155 1123 A-408 750 A-639 deposit2 = account_number, balance, tuple_id (employee_info ) Database System Concepts - 5th Edition, Aug 22, 2005. 22.10 1 2 3 4 5 6 7 tuple_id 1 2 3 4 5 6 7 ©Silberschatz, Korth and Sudarshan Advantages of Fragmentation    Horizontal:  allows parallel processing on fragments of a relation  allows a relation to be split so that tuples are located where they are most frequently accessed Vertical:  allows tuples to be split so that each part of the tuple is stored where it is most frequently accessed  tuple-id attribute allows efficient joining of vertical fragments  allows parallel processing on a relation Vertical and horizontal fragmentation can be mixed.  Fragments may be successively fragmented to an arbitrary depth. Database System Concepts - 5th Edition, Aug 22, 2005. 22.11 ©Silberschatz, Korth and Sudarshan Data Transparency  Data transparency: Degree to which system user may remain unaware of the details of how and where the data items are stored in a distributed system  Consider transparency issues in relation to:  Fragmentation transparency  Replication transparency  Location transparency Database System Concepts - 5th Edition, Aug 22, 2005. 22.12 ©Silberschatz, Korth and Sudarshan Naming of Data Items - Criteria 1. Every data item must have a system-wide unique name. 2. It should be possible to find the location of data items efficiently. 3. It should be possible to change the location of data items transparently. 4. Each site should be able to create new data items autonomously. Database System Concepts - 5th Edition, Aug 22, 2005. 22.13 ©Silberschatz, Korth and Sudarshan Centralized Scheme - Name Server   Structure:  name server assigns all names  each site maintains a record of local data items  sites ask name server to locate non-local data items Advantages:   satisfies naming criteria 1-3 Disadvantages:  does not satisfy naming criterion 4  name server is a potential performance bottleneck  name server is a single point of failure Database System Concepts - 5th Edition, Aug 22, 2005. 22.14 ©Silberschatz, Korth and Sudarshan Use of Aliases  Alternative to centralized scheme: each site prefixes its own site identifier to any name that it generates i.e., site 17.account.  Fulfills having a unique identifier, and avoids problems associated with central control.  However, fails to achieve network transparency.  Solution: Create a set of aliases for data items; Store the mapping of aliases to the real names at each site.  The user can be unaware of the physical location of a data item, and is unaffected if the data item is moved from one site to another. Database System Concepts - 5th Edition, Aug 22, 2005. 22.15 ©Silberschatz, Korth and Sudarshan Distributed Transactions  Transaction may access data at several sites.  Each site has a local transaction manager responsible for:   Maintaining a log for recovery purposes  Participating in coordinating the concurrent execution of the transactions executing at that site. Each site has a transaction coordinator, which is responsible for:  Starting the execution of transactions that originate at the site.  Distributing subtransactions at appropriate sites for execution.  Coordinating the termination of each transaction that originates at the site, which may result in the transaction being committed at all sites or aborted at all sites. Database System Concepts - 5th Edition, Aug 22, 2005. 22.16 ©Silberschatz, Korth and Sudarshan Transaction System Architecture Database System Concepts - 5th Edition, Aug 22, 2005. 22.17 ©Silberschatz, Korth and Sudarshan System Failure Modes  Failures unique to distributed systems:  Failure of a site.  Loss of massages   Failure of a communication link   Handled by network transmission control protocols such as TCP-IP Handled by network protocols, by routing messages via alternative links Network partition  A network is said to be partitioned when it has been split into two or more subsystems that lack any connection between them – Note: a subsystem may consist of a single node  Network partitioning and site failures are generally indistinguishable. Database System Concepts - 5th Edition, Aug 22, 2005. 22.18 ©Silberschatz, Korth and Sudarshan Commit Protocols  Commit protocols are used to ensure atomicity across sites  a transaction which executes at multiple sites must either be committed at all the sites, or aborted at all the sites.  not acceptable to have a transaction committed at one site and aborted at another  The two-phase commit (2PC) protocol is widely used  The three-phase commit (3PC) protocol is more complicated and more expensive, but avoids some drawbacks of two-phase commit protocol. This protocol is not used in practice. Database System Concepts - 5th Edition, Aug 22, 2005. 22.19 ©Silberschatz, Korth and Sudarshan Two Phase Commit Protocol (2PC)  Assumes fail-stop model – failed sites simply stop working, and do not cause any other harm, such as sending incorrect messages to other sites.  Execution of the protocol is initiated by the coordinator after the last step of the transaction has been reached.  The protocol involves all the local sites at which the transaction executed  Let T be a transaction initiated at site Si, and let the transaction coordinator at Si be Ci Database System Concepts - 5th Edition, Aug 22, 2005. 22.20 ©Silberschatz, Korth and Sudarshan Phase 1: Obtaining a Decision   Coordinator asks all participants to prepare to commit transaction Ti.  Ci adds the records <prepare T> to the log and forces log to stable storage  sends prepare T messages to all sites at which T executed Upon receiving message, transaction manager at site determines if it can commit the transaction  if not, add a record <no T> to the log and send abort T message to Ci  if the transaction can be committed, then:  add the record <ready  force T> to the log all records for T to stable storage  send ready T message to Ci Database System Concepts - 5th Edition, Aug 22, 2005. 22.21 ©Silberschatz, Korth and Sudarshan Phase 2: Recording the Decision  T can be committed of Ci received a ready T message from all the participating sites: otherwise T must be aborted.  Coordinator adds a decision record, <commit T> or <abort T>, to the log and forces record onto stable storage. Once the record stable storage it is irrevocable (even if failures occur)  Coordinator sends a message to each participant informing it of the decision (commit or abort)  Participants take appropriate action locally. Database System Concepts - 5th Edition, Aug 22, 2005. 22.22 ©Silberschatz, Korth and Sudarshan Handling of Failures - Site Failure When site Si recovers, it examines its log to determine the fate of transactions active at the time of the failure.  Log contain <commit T> record: site executes redo (T)  Log contains <abort T> record: site executes undo (T)  Log contains <ready T> record: site must consult Ci to determine the fate of T.   If T committed, redo (T)  If T aborted, undo (T) The log contains no control records concerning T replies that Sk failed before responding to the prepare T message from Ci  since the failure of Sk precludes the sending of such a response C1 must abort T  Sk must execute undo (T) Database System Concepts - 5th Edition, Aug 22, 2005. 22.23 ©Silberschatz, Korth and Sudarshan Handling of Failures- Coordinator Failure   If coordinator fails while the commit protocol for T is executing then participating sites must decide on T’s fate: 1. If an active site contains a <commit T> record in its log, then T must be committed. 2. If an active site contains an <abort T> record in its log, then T must be aborted. 3. If some active participating site does not contain a <ready T> record in its log, then the failed coordinator Ci cannot have decided to commit T. Can therefore abort T. 4. If none of the above cases holds, then all active sites must have a <ready T> record in their logs, but no additional control records (such as <abort T> of <commit T>). In this case active sites must wait for Ci to recover, to find decision. Blocking problem : active sites may have to wait for failed coordinator to recover. Database System Concepts - 5th Edition, Aug 22, 2005. 22.24 ©Silberschatz, Korth and Sudarshan Handling of Failures - Network Partition  If the coordinator and all its participants remain in one partition, the failure has no effect on the commit protocol.  If the coordinator and its participants belong to several partitions:  Sites that are not in the partition containing the coordinator think the coordinator has failed, and execute the protocol to deal with failure of the coordinator.   No harm results, but sites may still have to wait for decision from coordinator. The coordinator and the sites are in the same partition as the coordinator think that the sites in the other partition have failed, and follow the usual commit protocol.  Again, no harm results Database System Concepts - 5th Edition, Aug 22, 2005. 22.25 ©Silberschatz, Korth and Sudarshan Recovery and Concurrency Control  In-doubt transactions have a <ready T>, but neither a <commit T>, nor an <abort T> log record.  The recovering site must determine the commit-abort status of such transactions by contacting other sites; this can slow and potentially block recovery.  Recovery algorithms can note lock information in the log.   Instead of <ready T>, write out <ready T, L> L = list of locks held by T when the log is written (read locks can be omitted).  For every in-doubt transaction T, all the locks noted in the <ready T, L> log record are reacquired. After lock reacquisition, transaction processing can resume; the commit or rollback of in-doubt transactions is performed concurrently with the execution of new transactions. Database System Concepts - 5th Edition, Aug 22, 2005. 22.26 ©Silberschatz, Korth and Sudarshan Alternative Models of Transaction Processing   Notion of a single transaction spanning multiple sites is inappropriate for many applications  E.g. transaction crossing an organizational boundary  No organization would like to permit an externally initiated transaction to block local transactions for an indeterminate period Alternative models carry out transactions by sending messages  Code to handle messages must be carefully designed to ensure atomicity and durability properties for updates   Isolation cannot be guaranteed, in that intermediate stages are visible, but code must ensure no inconsistent states result due to concurrency Persistent messaging systems are systems that provide transactional properties to messages  Messages are guaranteed to be delivered exactly once  Will discuss implementation techniques later Database System Concepts - 5th Edition, Aug 22, 2005. 22.27 ©Silberschatz, Korth and Sudarshan Alternative Models (Cont.)  Motivating example: funds transfer between two banks  Two phase commit would have the potential to block updates on the accounts involved in funds transfer  Alternative solution:    Debit money from source account and send a message to other site  Site receives message and credits destination account Messaging has long been used for distributed transactions (even before computers were invented!) Atomicity issue  once transaction sending a message is committed, message must guaranteed to be delivered   Guarantee as long as destination site is up and reachable, code to handle undeliverable messages must also be available – e.g. credit money back to source account. If sending transaction aborts, message must not be sent Database System Concepts - 5th Edition, Aug 22, 2005. 22.28 ©Silberschatz, Korth and Sudarshan Error Conditions with Persistent Messaging  Code to handle messages has to take care of variety of failure situations (even assuming guaranteed message delivery)  E.g. if destination account does not exist, failure message must be sent back to source site  When failure message is received from destination site, or destination site itself does not exist, money must be deposited back in source account  Problem if source account has been closed – get humans to take care of problem  User code executing transaction processing using 2PC does not have to deal with such failures  There are many situations where extra effort of error handling is worth the benefit of absence of blocking  E.g. pretty much all transactions across organizations Database System Concepts - 5th Edition, Aug 22, 2005. 22.29 ©Silberschatz, Korth and Sudarshan Persistent Messaging and Workflows  Workflows provide a general model of transactional processing involving multiple sites and possibly human processing of certain steps  E.g. when a bank receives a loan application, it may need to  Contact external credit-checking agencies  Get approvals of one or more managers and then respond to the loan application  We study workflows in Chapter 25  Persistent messaging forms the underlying infrastructure for workflows in a distributed environment Database System Concepts - 5th Edition, Aug 22, 2005. 22.30 ©Silberschatz, Korth and Sudarshan Concurrency Control  Modify concurrency control schemes for use in distributed environment.  We assume that each site participates in the execution of a commit protocol to ensure global transaction automicity.  We assume all replicas of any item are updated  Will see how to relax this in case of site failures later Database System Concepts - 5th Edition, Aug 22, 2005. 22.31 ©Silberschatz, Korth and Sudarshan Single-Lock-Manager Approach  System maintains a single lock manager that resides in a single chosen site, say Si  When a transaction needs to lock a data item, it sends a lock request to S i and lock manager determines whether the lock can be granted immediately  If yes, lock manager sends a message to the site which initiated the request  If no, request is delayed until it can be granted, at which time a message is sent to the initiating site Database System Concepts - 5th Edition, Aug 22, 2005. 22.32 ©Silberschatz, Korth and Sudarshan Single-Lock-Manager Approach (Cont.)  The transaction can read the data item from any one of the sites at which a replica of the data item resides.  Writes must be performed on all replicas of a data item  Advantages of scheme:   Simple implementation  Simple deadlock handling Disadvantages of scheme are:  Bottleneck: lock manager site becomes a bottleneck  Vulnerability: system is vulnerable to lock manager site failure. Database System Concepts - 5th Edition, Aug 22, 2005. 22.33 ©Silberschatz, Korth and Sudarshan Distributed Lock Manager  In this approach, functionality of locking is implemented by lock managers at each site  Lock managers control access to local data items  But special protocols may be used for replicas  Advantage: work is distributed and can be made robust to failures  Disadvantage: deadlock detection is more complicated  Lock managers cooperate for deadlock detection   More on this later Several variants of this approach  Primary copy  Majority protocol  Biased protocol  Quorum consensus Database System Concepts - 5th Edition, Aug 22, 2005. 22.34 ©Silberschatz, Korth and Sudarshan Primary Copy   Choose one replica of data item to be the primary copy.  Site containing the replica is called the primary site for that data item  Different data items can have different primary sites When a transaction needs to lock a data item Q, it requests a lock at the primary site of Q.   Benefit   Implicitly gets lock on all replicas of the data item Concurrency control for replicated data handled similarly to unreplicated data - simple implementation. Drawback  If the primary site of Q fails, Q is inaccessible even though other sites containing a replica may be accessible. Database System Concepts - 5th Edition, Aug 22, 2005. 22.35 ©Silberschatz, Korth and Sudarshan Majority Protocol  Local lock manager at each site administers lock and unlock requests for data items stored at that site.  When a transaction wishes to lock an unreplicated data item Q residing at site Si, a message is sent to Si ‘s lock manager.  If Q is locked in an incompatible mode, then the request is delayed until it can be granted.  When the lock request can be granted, the lock manager sends a message back to the initiator indicating that the lock request has been granted. Database System Concepts - 5th Edition, Aug 22, 2005. 22.36 ©Silberschatz, Korth and Sudarshan Majority Protocol (Cont.)   In case of replicated data  If Q is replicated at n sites, then a lock request message must be sent to more than half of the n sites in which Q is stored.  The transaction does not operate on Q until it has obtained a lock on a majority of the replicas of Q.  When writing the data item, transaction performs writes on all replicas. Benefit  Can be used even when some sites are unavailable   details on how handle writes in the presence of site failure later Drawback  Requires 2(n/2 + 1) messages for handling lock requests, and ( n/2 + 1) messages for handling unlock requests.  Potential for deadlock even with single item - e.g., each of 3 transactions may have locks on 1/3rd of the replicas of a data. Database System Concepts - 5th Edition, Aug 22, 2005. 22.37 ©Silberschatz, Korth and Sudarshan Biased Protocol  Local lock manager at each site as in majority protocol, however, requests for shared locks are handled differently than requests for exclusive locks.  Shared locks. When a transaction needs to lock data item Q, it simply requests a lock on Q from the lock manager at one site containing a replica of Q.  Exclusive locks. When transaction needs to lock data item Q, it requests a lock on Q from the lock manager at all sites containing a replica of Q.  Advantage - imposes less overhead on read operations.  Disadvantage - additional overhead on writes Database System Concepts - 5th Edition, Aug 22, 2005. 22.38 ©Silberschatz, Korth and Sudarshan Quorum Consensus Protocol  A generalization of both majority and biased protocols  Each site is assigned a weight.   Let S be the total of all site weights Choose two values read quorum Qr and write quorum Qw  Such that Q r + Qw > S and 2 * Qw > S  Quorums can be chosen (and S computed) separately for each item  Each read must lock enough replicas that the sum of the site weights is >= Q r  Each write must lock enough replicas that the sum of the site weights is >= Q w  For now we assume all replicas are written  Extensions to allow some sites to be unavailable described later Database System Concepts - 5th Edition, Aug 22, 2005. 22.39 ©Silberschatz, Korth and Sudarshan Timestamping  Timestamp based concurrency-control protocols can be used in distributed systems  Each transaction must be given a unique timestamp  Main problem: how to generate a timestamp in a distributed fashion  Each site generates a unique local timestamp using either a logical counter or the local clock.  Global unique timestamp is obtained by concatenating the unique local timestamp with the unique identifier. Database System Concepts - 5th Edition, Aug 22, 2005. 22.40 ©Silberschatz, Korth and Sudarshan Timestamping (Cont.)   A site with a slow clock will assign smaller timestamps  Still logically correct: serializability not affected  But: “disadvantages” transactions To fix this problem  Define within each site Si a logical clock (LCi), which generates the unique local timestamp  Require that Si advance its logical clock whenever a request is received from a transaction Ti with timestamp < x,y> and x is greater that the current value of LCi.  In this case, site Si advances its logical clock to the value x + 1. Database System Concepts - 5th Edition, Aug 22, 2005. 22.41 ©Silberschatz, Korth and Sudarshan Replication with Weak Consistency  Many commercial databases support replication of data with weak degrees of consistency (I.e., without a guarantee of serializabiliy)  E.g.: master-slave replication: updates are performed at a single “master” site, and propagated to “slave” sites.   Propagation is not part of the update transaction: its is decoupled  May be immediately after transaction commits  May be periodic Data may only be read at slave sites, not updated   Particularly useful for distributing information   No need to obtain locks at any remote site E.g. from central office to branch-office Also useful for running read-only queries offline from the main database Database System Concepts - 5th Edition, Aug 22, 2005. 22.42 ©Silberschatz, Korth and Sudarshan Replication with Weak Consistency (Cont.)  Replicas should see a transaction-consistent snapshot of the database   That is, a state of the database reflecting all effects of all transactions up to some point in the serialization order, and no effects of any later transactions. E.g. Oracle provides a create snapshot statement to create a snapshot of a relation or a set of relations at a remote site  snapshot refresh either by recomputation or by incremental update  Automatic refresh (continuous or periodic) or manual refresh Database System Concepts - 5th Edition, Aug 22, 2005. 22.43 ©Silberschatz, Korth and Sudarshan Multimaster and Lazy Replication  With multimaster replication (also called update-anywhere replication) updates are permitted at any replica, and are automatically propagated to all replicas  Basic model in distributed databases, where transactions are unaware of the details of replication, and database system propagates updates as part of the same transaction   Coupled with 2 phase commit Many systems support lazy propagation where updates are transmitted after transaction commits  Allows updates to occur even if some sites are disconnected from the network, but at the cost of consistency Database System Concepts - 5th Edition, Aug 22, 2005. 22.44 ©Silberschatz, Korth and Sudarshan Deadlock Handling Consider the following two transactions and history, with item X and transaction T1 at site 1, and item Y and transaction T2 at site 2: T1: T2: write (X) write (Y) X-lock on X write (X) write (Y) write (X) X-lock on Y write (Y) wait for X-lock on X Wait for X-lock on Y Result: deadlock which cannot be detected locally at either site Database System Concepts - 5th Edition, Aug 22, 2005. 22.45 ©Silberschatz, Korth and Sudarshan Centralized Approach    A global wait-for graph is constructed and maintained in a single site; the deadlock-detection coordinator  Real graph: Real, but unknown, state of the system.  Constructed graph:Approximation generated by the controller during the execution of its algorithm . the global wait-for graph can be constructed when:  a new edge is inserted in or removed from one of the local wait-for graphs.  a number of changes have occurred in a local wait-for graph.  the coordinator needs to invoke cycle-detection. If the coordinator finds a cycle, it selects a victim and notifies all sites. The sites roll back the victim transaction. Database System Concepts - 5th Edition, Aug 22, 2005. 22.46 ©Silberschatz, Korth and Sudarshan Local and Global Wait-For Graphs Local Global Database System Concepts - 5th Edition, Aug 22, 2005. 22.47 ©Silberschatz, Korth and Sudarshan Example Wait-For Graph for False Cycles Initial state: Database System Concepts - 5th Edition, Aug 22, 2005. 22.48 ©Silberschatz, Korth and Sudarshan False Cycles (Cont.)  Suppose that starting from the state shown in figure, 1. T2 releases resources at S1  resulting in a message remove T1  T2 message from the Transaction Manager at site S1 to the coordinator) 2. And then T2 requests a resource held by T3 at site S2   Suppose further that the insert message reaches before the delete message   resulting in a message insert T2  T3 from S2 to the coordinator this can happen due to network delays The coordinator would then find a false cycle T1  T2  T3  T1  The false cycle above never existed in reality.  False cycles cannot occur if two-phase locking is used. Database System Concepts - 5th Edition, Aug 22, 2005. 22.49 ©Silberschatz, Korth and Sudarshan Unnecessary Rollbacks  Unnecessary rollbacks may result when deadlock has indeed occurred and a victim has been picked, and meanwhile one of the transactions was aborted for reasons unrelated to the deadlock.  Unnecessary rollbacks can result from false cycles in the global wait-for graph; however, likelihood of false cycles is low. Database System Concepts - 5th Edition, Aug 22, 2005. 22.50 ©Silberschatz, Korth and Sudarshan Availability  High availability: time for which system is not fully usable should be extremely low (e.g. 99.99% availability)  Robustness: ability of system to function spite of failures of components  Failures are more likely in large distributed systems  To be robust, a distributed system must   Detect failures  Reconfigure the system so computation may continue  Recovery/reintegration when a site or link is repaired Failure detection: distinguishing link failure from site failure is hard  (partial) solution: have multiple links, multiple link failure is likely a site failure Database System Concepts - 5th Edition, Aug 22, 2005. 22.51 ©Silberschatz, Korth and Sudarshan Reconfiguration  Reconfiguration:   Abort all transactions that were active at a failed site  Making them wait could interfere with other transactions since they may hold locks on other sites  However, in case only some replicas of a data item failed, it may be possible to continue transactions that had accessed data at a failed site (more on this later) If replicated data items were at failed site, update system catalog to remove them from the list of replicas.   This should be reversed when failed site recovers, but additional care needs to be taken to bring values up to date If a failed site was a central server for some subsystem, an election must be held to determine the new server  E.g. name server, concurrency coordinator, global deadlock detector Database System Concepts - 5th Edition, Aug 22, 2005. 22.52 ©Silberschatz, Korth and Sudarshan Reconfiguration (Cont.)  Since network partition may not be distinguishable from site failure, the following situations must be avoided  Two ore more central servers elected in distinct partitions  More than one partition updates a replicated data item  Updates must be able to continue even if some sites are down  Solution: majority based approach  Alternative of “read one write all available” is tantalizing but causes problems Database System Concepts - 5th Edition, Aug 22, 2005. 22.53 ©Silberschatz, Korth and Sudarshan Majority-Based Approach  The majority protocol for distributed concurrency control can be modified to work even if some sites are unavailable  Each replica of each item has a version number which is updated when the replica is updated, as outlined below  A lock request is sent to at least ½ the sites at which item replicas are stored and operation continues only when a lock is obtained on a majority of the sites  Read operations look at all replicas locked, and read the value from the replica with largest version number  May write this value and version number back to replicas with lower version numbers (no need to obtain locks on all replicas for this task) Database System Concepts - 5th Edition, Aug 22, 2005. 22.54 ©Silberschatz, Korth and Sudarshan Majority-Based Approach  Majority protocol (Cont.)    Write operations  find highest version number like reads, and set new version number to old highest version + 1  Writes are then performed on all locked replicas and version number on these replicas is set to new version number Failures (network and site) cause no problems as long as  Sites at commit contain a majority of replicas of any updated data items  During reads a majority of replicas are available to find version numbers  Subject to above, 2 phase commit can be used to update replicas  Note: reads are guaranteed to see latest version of data item  Reintegration is trivial: nothing needs to be done Quorum consensus algorithm can be similarly extended Database System Concepts - 5th Edition, Aug 22, 2005. 22.55 ©Silberschatz, Korth and Sudarshan Read One Write All (Available)  Biased protocol is a special case of quorum consensus   Allows reads to read any one replica but updates require all replicas to be available at commit time (called read one write all) Read one write all available (ignoring failed sites) is attractive, but incorrect  If failed link may come back up, without a disconnected site ever being aware that it was disconnected  The site then has old values, and a read from that site would return an incorrect value  If site was aware of failure reintegration could have been performed, but no way to guarantee this  With network partitioning, sites in each partition may update same item concurrently  believing sites in other partitions have all failed Database System Concepts - 5th Edition, Aug 22, 2005. 22.56 ©Silberschatz, Korth and Sudarshan Site Reintegration  When failed site recovers, it must catch up with all updates that it missed while it was down  Problem: updates may be happening to items whose replica is stored at the site while the site is recovering  Solution 1: halt all updates on system while reintegrating a site   Unacceptable disruption Solution 2: lock all replicas of all data items at the site, update to latest version, then release locks  Other solutions with better concurrency also available Database System Concepts - 5th Edition, Aug 22, 2005. 22.57 ©Silberschatz, Korth and Sudarshan Comparison with Remote Backup  Remote backup (hot spare) systems (Section 17.10) are also designed to provide high availability  Remote backup systems are simpler and have lower overhead   All actions performed at a single site, and only log records shipped  No need for distributed concurrency control, or 2 phase commit Using distributed databases with replicas of data items can provide higher availability by having multiple (> 2) replicas and using the majority protocol  Also avoid failure detection and switchover time associated with remote backup systems Database System Concepts - 5th Edition, Aug 22, 2005. 22.58 ©Silberschatz, Korth and Sudarshan Coordinator Selection   Backup coordinators  site which maintains enough information locally to assume the role of coordinator if the actual coordinator fails  executes the same algorithms and maintains the same internal state information as the actual coordinator fails executes state information as the actual coordinator  allows fast recovery from coordinator failure but involves overhead during normal processing. Election algorithms  used to elect a new coordinator in case of failures  Example: Bully Algorithm - applicable to systems where every site can send a message to every other site. Database System Concepts - 5th Edition, Aug 22, 2005. 22.59 ©Silberschatz, Korth and Sudarshan Bully Algorithm  If site Si sends a request that is not answered by the coordinator within a time interval T, assume that the coordinator has failed Si tries to elect itself as the new coordinator.  Si sends an election message to every site with a higher identification number, S i then waits for any of these processes to answer within T.  If no response within T, assume that all sites with number greater than i have failed, Si elects itself the new coordinator.  If answer is received Si begins time interval T’, waiting to receive a message that a site with a higher identification number has been elected. Database System Concepts - 5th Edition, Aug 22, 2005. 22.60 ©Silberschatz, Korth and Sudarshan Bully Algorithm (Cont.)  If no message is sent within T’, assume the site with a higher number has failed; Si restarts the algorithm.  After a failed site recovers, it immediately begins execution of the same algorithm.  If there are no active sites with higher numbers, the recovered site forces all processes with lower numbers to let it become the coordinator site, even if there is a currently active coordinator with a lower number. Database System Concepts - 5th Edition, Aug 22, 2005. 22.61 ©Silberschatz, Korth and Sudarshan Distributed Query Processing  For centralized systems, the primary criterion for measuring the cost of a particular strategy is the number of disk accesses.  In a distributed system, other issues must be taken into account:  The cost of a data transmission over the network.  The potential gain in performance from having several sites process parts of the query in parallel. Database System Concepts - 5th Edition, Aug 22, 2005. 22.62 ©Silberschatz, Korth and Sudarshan Query Transformation   Translating algebraic queries on fragments.  It must be possible to construct relation r from its fragments  Replace relation r by the expression to construct relation r from its fragments Consider the horizontal fragmentation of the account relation into account1 =  branch_name = “Hillside” (account ) account2 =  branch_name = “Valleyview” (account )  The query  branch_name = “Hillside” (account ) becomes  branch_name = “Hillside” (account1  account2) which is optimized into  branch_name = “Hillside” (account1)   branch_name = “Hillside” (account2) Database System Concepts - 5th Edition, Aug 22, 2005. 22.63 ©Silberschatz, Korth and Sudarshan Example Query (Cont.)  Since account1 has only tuples pertaining to the Hillside branch, we can eliminate the selection operation.  Apply the definition of account2 to obtain  branch_name = “Hillside” ( branch_name = “Valleyview” (account )  This expression is the empty set regardless of the contents of the account relation.  Final strategy is for the Hillside site to return account1 as the result of the query. Database System Concepts - 5th Edition, Aug 22, 2005. 22.64 ©Silberschatz, Korth and Sudarshan Simple Join Processing  Consider the following relational algebra expression in which the three relations are neither replicated nor fragmented account depositor branch  account is stored at site S1  depositor at S2  branch at S3  For a query issued at site SI, the system needs to produce the result at site SI Database System Concepts - 5th Edition, Aug 22, 2005. 22.65 ©Silberschatz, Korth and Sudarshan Possible Query Processing Strategies  Ship copies of all three relations to site SI and choose a strategy for processing the entire locally at site SI.  Ship a copy of the account relation to site S2 and compute temp1 = account depositor at S2. Ship temp1 from S2 to S3, and compute temp2 = temp1 branch at S3. Ship the result temp2 to SI.  Devise similar strategies, exchanging the roles S1, S2, S3  Must consider following factors:  amount of data being shipped  cost of transmitting a data block between sites  relative processing speed at each site Database System Concepts - 5th Edition, Aug 22, 2005. 22.66 ©Silberschatz, Korth and Sudarshan Semijoin Strategy  Let r1 be a relation with schema R1 stores at site S1 Let r2 be a relation with schema R2 stores at site S2  Evaluate the expression r1 r2 and obtain the result at S1. 1. Compute temp1  R1  R2 (r1) at S1.  2. Ship temp1 from S1 to S2.  3. Compute temp2  r2  4. Ship temp2 from S2 to S1.  5. Compute r1 temp1 at S2 temp2 at S1. This is the same as r1 Database System Concepts - 5th Edition, Aug 22, 2005. 22.67 r2. ©Silberschatz, Korth and Sudarshan Formal Definition  The semijoin of r1 with r2, is denoted by: r1 r2  it is defined by:  R1 (r1 r2)  Thus, r1 r2 selects those tuples of r1 that contributed to r1  In step 3 above, temp2=r2  For joins of several relations, the above strategy can be extended to a series of semijoin steps. Database System Concepts - 5th Edition, Aug 22, 2005. r2. r1. 22.68 ©Silberschatz, Korth and Sudarshan Join Strategies that Exploit Parallelism  Consider r1 r2 r3 r4 where relation ri is stored at site Si. The result must be presented at site S1.  r1 is shipped to S2 and r1 S4 and r3 r2 is computed at S2: simultaneously r3 is shipped to r4 is computed at S4  S2 ships tuples of (r1 S4 ships tuples of (r3  Once tuples of (r1 r2) to S1 as they produced; r4) to S1 r2) and (r3 r4) arrive at S1 (r1 computed in parallel with the computation of (r1 (r3 r2) (r3 r4) is r2) at S2 and the computation of r4) at S4. Database System Concepts - 5th Edition, Aug 22, 2005. 22.69 ©Silberschatz, Korth and Sudarshan Heterogeneous Distributed Databases  Many database applications require data from a variety of preexisting databases located in a heterogeneous collection of hardware and software platforms  Data models may differ (hierarchical, relational , etc.)  Transaction commit protocols may be incompatible  Concurrency control may be based on different techniques (locking, timestamping, etc.)  System-level details almost certainly are totally incompatible.  A multidatabase system is a software layer on top of existing database systems, which is designed to manipulate information in heterogeneous databases  Creates an illusion of logical database integration without any physical database integration Database System Concepts - 5th Edition, Aug 22, 2005. 22.70 ©Silberschatz, Korth and Sudarshan Advantages  Preservation of investment in existing  hardware  system software  Applications  Local autonomy and administrative control  Allows use of special-purpose DBMSs  Step towards a unified homogeneous DBMS  Full integration into a homogeneous DBMS faces  Technical difficulties and cost of conversion  Organizational/political difficulties – Organizations do not want to give up control on their data – Local databases wish to retain a great deal of autonomy Database System Concepts - 5th Edition, Aug 22, 2005. 22.71 ©Silberschatz, Korth and Sudarshan Unified View of Data  Agreement on a common data model    Typically the relational model Agreement on a common conceptual schema  Different names for same relation/attribute  Same relation/attribute name means different things Agreement on a single representation of shared data  E.g. data types, precision,  Character sets  ASCII vs EBCDIC  Sort order variations  Agreement on units of measure  Variations in names  E.g. Köln vs Cologne, Mumbai vs Bombay Database System Concepts - 5th Edition, Aug 22, 2005. 22.72 ©Silberschatz, Korth and Sudarshan Query Processing  Several issues in query processing in a heterogeneous database  Schema translation   Write a wrapper for each data source to translate data to a global schema  Wrappers must also translate updates on global schema to updates on local schema Limited query capabilities  Some data sources allow only restricted forms of selections    Queries have to be broken up and processed partly at the source and partly at a different site Removal of duplicate information when sites have overlapping information   E.g. web forms, flat file data sources Decide which sites to execute query Global query optimization Database System Concepts - 5th Edition, Aug 22, 2005. 22.73 ©Silberschatz, Korth and Sudarshan Mediator Systems  Mediator systems are systems that integrate multiple heterogeneous data sources by providing an integrated global view, and providing query facilities on global view  Unlike full fledged multidatabase systems, mediators generally do not bother about transaction processing  But the terms mediator and multidatabase are sometimes used interchangeably  The term virtual database is also used to refer to mediator/multidatabase systems Database System Concepts - 5th Edition, Aug 22, 2005. 22.74 ©Silberschatz, Korth and Sudarshan Directory Systems  Typical kinds of directory information  Employee information such as name, id, email, phone, office addr, ..  Even personal information to be accessed from multiple places   White pages  Entries organized by name or identifier    e.g. Web browser bookmarks Meant for forward lookup to find more about an entry Yellow pages  Entries organized by properties  For reverse lookup to find entries matching specific requirements When directories are to be accessed across an organization  Alternative 1: Web interface. Not great for programs  Alternative 2: Specialized directory access protocols  Coupled with specialized user interfaces Database System Concepts - 5th Edition, Aug 22, 2005. 22.75 ©Silberschatz, Korth and Sudarshan Directory Access Protocols  Most commonly used directory access protocol:  LDAP (Lightweight Directory Access Protocol)  Simplified from earlier X.500 protocol  Question: Why not use database protocols like ODBC/JDBC?  Answer:  Simplified protocols for a limited type of data access, evolved parallel to ODBC/JDBC  Provide a nice hierarchical naming mechanism similar to file system directories  Data can be partitioned amongst multiple servers for different parts of the hierarchy, yet give a single view to user – E.g. different servers for Bell Labs Murray Hill and Bell Labs Bangalore  Directories may use databases as storage mechanism Database System Concepts - 5th Edition, Aug 22, 2005. 22.76 ©Silberschatz, Korth and Sudarshan LDAP: Lightweight Directory Access Protocol  LDAP Data Model  Data Manipulation  Distributed Directory Trees Database System Concepts - 5th Edition, Aug 22, 2005. 22.77 ©Silberschatz, Korth and Sudarshan LDAP Data Model  LDAP directories store entries  Entries are similar to objects  Each entry must have unique distinguished name (DN)  DN made up of a sequence of relative distinguished names (RDNs)  E.g. of a DN  cn=Silberschatz, ou-Bell Labs, o=Lucent, c=USA  Standard RDNs (can be specified as part of schema)   cn: common name ou: organizational unit  o: organization c: country Similar to paths in a file system but written in reverse direction Database System Concepts - 5th Edition, Aug 22, 2005. 22.78 ©Silberschatz, Korth and Sudarshan LDAP Data Model (Cont.)   Entries can have attributes  Attributes are multi-valued by default  LDAP has several built-in types  Binary, string, time types  Tel: telephone number PostalAddress: postal address LDAP allows definition of object classes  Object classes specify attribute names and types  Can use inheritance to define object classes  Entry can be specified to be of one or more object classes  No need to have single most-specific type Database System Concepts - 5th Edition, Aug 22, 2005. 22.79 ©Silberschatz, Korth and Sudarshan LDAP Data Model (cont.)  Entries organized into a directory information tree according to their DNs  Leaf level usually represent specific objects  Internal node entries represent objects such as organizational units, organizations or countries  Children of a node inherit the DN of the parent, and add on RDNs   E.g. internal node with DN c=USA – Children nodes have DN starting with c=USA and further RDNs such as o or ou  DN of an entry can be generated by traversing path from root Leaf level can be an alias pointing to another entry  Entries can thus have more than one DN – E.g. person in more than one organizational unit Database System Concepts - 5th Edition, Aug 22, 2005. 22.80 ©Silberschatz, Korth and Sudarshan LDAP Data Manipulation  Unlike SQL, LDAP does not define DDL or DML  Instead, it defines a network protocol for DDL and DML  Users use an API or vendor specific front ends  LDAP also defines a file format   LDAP Data Interchange Format (LDIF) Querying mechanism is very simple: only selection & projection Database System Concepts - 5th Edition, Aug 22, 2005. 22.81 ©Silberschatz, Korth and Sudarshan LDAP Queries  LDAP query must specify  Base: a node in the DIT from where search is to start  A search condition  Boolean combination of conditions on attributes of entries – Equality, wild-cards and approximate equality supported  A scope  Just the base, the base and its children, or the entire subtree from the base  Attributes to be returned  Limits on number of results and on resource consumption  May also specify whether to automatically dereference aliases  LDAP URLs are one way of specifying query  LDAP API is another alternative Database System Concepts - 5th Edition, Aug 22, 2005. 22.82 ©Silberschatz, Korth and Sudarshan LDAP URLs  First part of URL specifis server and DN of base   ldap:://aura.research.bell-labs.com/o=Lucent,c=USA Optional further parts separated by ? symbol  ldap:://aura.research.bell-labs.com/o=Lucent,c=USA??sub?cn=Korth  Optional parts specify 1. attributes to return (empty means all) 2. Scope (sub indicates entire subtree) 3. Search condition (cn=Korth) Database System Concepts - 5th Edition, Aug 22, 2005. 22.83 ©Silberschatz, Korth and Sudarshan C Code using LDAP API #include <stdio.h> #include <ldap.h> main( ) { LDAP *ld; LDAPMessage *res, *entry; char *dn, *attr, *attrList [ ] = {“telephoneNumber”, NULL}; BerElement *ptr; int vals, i; // Open a connection to server ld = ldap_open(“aura.research.bell-labs.com”, LDAP_PORT); ldap_simple_bind(ld, “avi”, “avi-passwd”); … actual query (next slide) … ldap_unbind(ld); } Database System Concepts - 5th Edition, Aug 22, 2005. 22.84 ©Silberschatz, Korth and Sudarshan C Code using LDAP API (Cont.) ldap_search_s(ld, “o=Lucent, c=USA”, LDAP_SCOPE_SUBTREE, “cn=Korth”, attrList, /* attrsonly*/ 0, &res); /*attrsonly = 1 => return only schema not actual results*/ printf(“found%d entries”, ldap_count_entries(ld, res)); for (entry=ldap_first_entry(ld, res); entry != NULL; entry=ldap_next_entry(id, entry)) { dn = ldap_get_dn(ld, entry); printf(“dn: %s”, dn); /* dn: DN of matching entry */ ldap_memfree(dn); for(attr = ldap_first_attribute(ld, entry, &ptr); attr != NULL; attr = ldap_next_attribute(ld, entry, ptr)) { // for each attribute printf(“%s:”, attr); // print name of attribute vals = ldap_get_values(ld, entry, attr); for (i = 0; vals[i] != NULL; i ++) printf(“%s”, vals[i]); // since attrs can be multivalued ldap_value_free(vals); } } ldap_msgfree(res); Database System Concepts - 5th Edition, Aug 22, 2005. 22.85 ©Silberschatz, Korth and Sudarshan LDAP API (Cont.)  LDAP API also has functions to create, update and delete entries  Each function call behaves as a separate transaction  LDAP does not support atomicity of updates Database System Concepts - 5th Edition, Aug 22, 2005. 22.86 ©Silberschatz, Korth and Sudarshan Distributed Directory Trees  Organizational information may be split into multiple directory information trees  Suffix of a DIT gives RDN to be tagged onto to all entries to get an overall DN   E.g. two DITs, one with suffix o=Lucent, c=USA and another with suffix o=Lucent, c=India  Organizations often split up DITs based on geographical location or by organizational structure  Many LDAP implementations support replication (master-slave or multi-master replication) of DITs (not part of LDAP 3 standard) A node in a DIT may be a referral to a node in another DIT  E.g. Ou= Bell Labs may have a separate DIT, and DIT for o=Lucent may have a leaf with ou=Bell Labs containing a referral to the Bell Labs DIT  Referalls are the key to integrating a distributed collection of directories  When a server gets a query reaching a referral node, it may either  Forward query to referred DIT and return answer to client, or  Give referral back to client, which transparently sends query to referred DIT (without user intervention) Database System Concepts - 5th Edition, Aug 22, 2005. 22.87 ©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 Three Phase Commit (3PC)   Assumptions:  No network partitioning  At any point, at least one site must be up.  At most K sites (participants as well as coordinator) can fail Phase 1: Obtaining Preliminary Decision: Identical to 2PC Phase 1.    Phase 2 of 2PC is split into 2 phases, Phase 2 and Phase 3 of 3PC  In phase 2 coordinator makes a decision as in 2PC (called the pre-commit decision) and records it in multiple (at least K) sites  In phase 3, coordinator sends commit/abort message to all participating sites, Under 3PC, knowledge of pre-commit decision can be used to commit despite coordinator failure   Every site is ready to commit if instructed to do so Avoids blocking problem as long as < K sites fail Drawbacks:  higher overheads  assumptions may not be satisfied in practice Database System Concepts - 5th Edition, Aug 22, 2005. 22.89 ©Silberschatz, Korth and Sudarshan Figure 22.3 Database System Concepts - 5th Edition, Aug 22, 2005. 22.90 ©Silberschatz, Korth and Sudarshan Figure 22.4 Database System Concepts - 5th Edition, Aug 22, 2005. 22.91 ©Silberschatz, Korth and Sudarshan Figure 22.5 Database System Concepts - 5th Edition, Aug 22, 2005. 22.92 ©Silberschatz, Korth and Sudarshan Figure 22.7 Database System Concepts - 5th Edition, Aug 22, 2005. 22.93 ©Silberschatz, Korth and Sudarshan