ASSIGNMENT OF INFORMATION TECHNOLOGY

SUBMITTED TO:

SUBMITTED BY:

Ms. KAMAL

VIPAN

DATA WAREHOUSE
A data warehouse is a repositor o! a" or#a"i$atio"%s e&e'tro"i'a&& stored data. Mea"s to retrie(e a"d a"a& $e data) to e*tra't) tra"s!or+ a"d &oad data) a"d to +a"a#e the data di'tio"ar are a&so 'o"sidered esse"tia& 'o+po"e"ts o! a data warehousi"# s ste+. Data warehousi"# i"'&udes ,usi"ess i"te&&i#e"'e too&s) too&s to e*tra't) tra"s!or+) a"d &oad data i"to the repositor ) a"d too&s to +a"a#e a"d retrie(e +etadata. According to Inmon, famous author for several data warehouse oo!s, "A data warehouse is a su #ect oriented, integrated, time variant, non volatile collection of data in su$$ort of management%s decision ma!ing $rocess&' Data warehousi"# arises i" a" or#a"i$atio"-s "eed !or re&ia,&e) 'o"so&idated) u"i.ue a"d i"te#rated reporti"# a"d a"a& sis o! its data) at di!!ere"t &e(e&s o! a##re#atio". The data warehousi"# 'o"su&ta"t is 'har#ed with +a/i"# the data appear 'o"siste"t) i"te#rated a"d 'o"so&idated despite the pro,&e+s i" the u"der& i"# sour'e s ste+s. The data warehousi"# 'o"su&ta"t a'hie(es this , e+p&o i"# di!!ere"t data warehousi"# te'h"i.ues) 'reati"# o"e or +ore "ew data repositories 0i.e. the data warehouse1 whose data +ode&0s1 support the "eeded reporti"# a"d a"a& sis. (e) develo$ments in earl) )ears of data warehousing were*

2345s 6 7e"era& Mi&&s a"d Dart+outh 8o&&e#e) i" a 9oi"t resear'h pro9e't) de(e&op the ter+s di+e"sio"s a"d !a'ts.:;< 23=5s 6 A8Nie&se" a"d I>I pro(ide di+e"sio"a& data +arts !or retai& sa&es.:;< 23?; 6 Teradata i"trodu'es a data,ase +a"a#e+e"t s ste+ spe'i!i'a&& desi#"ed !or de'isio" support. 23?? 6 Barr De(&i" a"d Pau& Murph pu,&ish the arti'&e A" ar'hite'ture !or a ,usi"ess a"d i"!or+atio" s ste+s i" IBM S ste+s @our"a& where the i"trodu'e the ter+ A,usi"ess data warehouseA. 2335 6 >ed Bri'/ S ste+s i"trodu'es >ed Bri'/ Barehouse) a data,ase +a"a#e+e"t s ste+ spe'i!i'a&& !or data warehousi"#. 2332 6 Pris+ So&utio"s i"trodu'es Pris+ Barehouse Ma"a#er) so!tware !or de(e&opi"# a data warehouse. 2332 6 Bi&& I"+o" pu,&ishes the ,oo/ Bui&di"# the Data Barehouse.

233C 6 The Data Barehousi"# I"stitute) a !orDpro!it or#a"i$atio" that pro+otes data warehousi"#) is !ou"ded. 2334 6 >a&ph Ki+,a&& pu,&ishes the ,oo/ The Data Barehouse Too&/it. 233= 6 Ora'&e ?) with support !or star .ueries) is re&eased. 233? 6 Mi'roso!t re&eases Mi'roso!t A"a& sis Ser(i'es 0the" OLAP Ser(i'es1 hea(i& uti&i$i"# data warehousi"# s'he+as.

E+am$le*
I" order to store data) o(er the ears) +a" app&i'atio" desi#"ers i" ea'h ,ra"'h ha(e +ade their i"di(idua& de'isio"s as to how a" app&i'atio" a"d data,ase shou&d ,e ,ui&t. So sour'e s ste+s wi&& ,e di!!ere"t i" "a+i"# 'o"(e"tio"s) (aria,&e +easure+e"ts) e"'odi"# stru'tures) a"d ph si'a& attri,utes o! data. 8o"sider a ,a"/ that has #ot se(era& ,ra"'hes i" se(era& 'ou"tries) has +i&&io"s o! 'usto+ers a"d the &i"es o! ,usi"ess o! the e"terprise are sa(i"#s) a"d &oa"s. The !o&&owi"# e*a+p&e e*p&ai"s how the data is i"te#rated !ro+ sour'e s ste+s to tar#et s ste+s. E+am$le of Source Data S)stem ,ame Attri ute ,ame -olumn ,ame Data t)$e .alues

8usto+er Sour'e App&i'atio" S ste+ 2 Date 8usto+er Sour'e App&i'atio" S ste+ F Date Sour'e App&i'atio" S ste+ ; Date

8USTOME>EAPPLI8ATIONEDATE NUME>I80?)51 2252F55C

8USTEAPPLI8ATIONEDATE APPLI8ATIONEDATE

DATE DATE

2252F55C 52NOVF55C

I" the a!ore+e"tio"ed e*a+p&e) attri,ute "a+e) 'o&u+" "a+e) data t pe a"d (a&ues are e"tire& di!!ere"t !ro+ o"e sour'e s ste+ to a"other. This i"'o"siste"' i" data 'a" ,e a(oided , i"te#rati"# the data i"to a data warehouse with #ood sta"dards. E+am$le of Target Data /Data Warehouse0 Target S)stem >e'ord G2 Attri ute ,ame -olumn ,ame Data t)$e .alues 5222F55C

8usto+er App&i'atio" 8USTOME>EAPPLI8ATIONEDATE DATE

Date >e'ord GF >e'ord G; 8usto+er App&i'atio" 8USTOME>EAPPLI8ATIONEDATE DATE Date 8usto+er App&i'atio" 8USTOME>EAPPLI8ATIONEDATE DATE Date 5222F55C 5222F55C

I" the a,o(e e*a+p&e o! tar#et data) attri,ute "a+es) 'o&u+" "a+es) a"d data t pes are 'o"siste"t throu#hout the tar#et s ste+. This is how data !ro+ (arious sour'e s ste+s is i"te#rated a"d a''urate& stored i"to the data warehouse.

AR-HITE-TURE

Data warehouse ar'hite'ture is pri+ari& ,ased o" the ,usi"ess pro'esses o! a ,usi"ess e"terprise ta/i"# i"to 'o"sideratio" the data 'o"so&idatio" a'ross the ,usi"ess e"terprise with ade.uate se'urit ) data +ode&i"# a"d or#a"i$atio") e*te"t o! .uer re.uire+e"ts) +eta data +a"a#e+e"t a"d app&i'atio") warehouse sta#i"# area p&a""i"# !or opti+u+ ,a"dwidth uti&i$atio" a"d !u&& te'h"o&o# i+p&e+e"tatio". The Data Barehouse Ar'hite'ture i"'&udes +a" !a'ets. So+e o! these are &isted as !o&&ows: Pro'ess Ar'hite'ture Data Mode& Ar'hite'ture Te'h"o&o# Ar'hite'ture

I"!or+atio" Ar'hite'ture >esour'e Ar'hite'ture

1rocess Architecture
Des'ri,es the "u+,er o! sta#es a"d how data is pro'essed to 'o"(ert raw H tra"sa'tio"a& data i"to i"!or+atio" !or e"d user usa#e. The data sta#i"# pro'ess i"'&udes three +ai" areas o! 'o"'er"s or su,D pro'esses !or p&a""i"# data warehouse ar'hite'ture "a+e& IE*tra'tJ) ITra"s!or+J a"d ILoadJ. These i"terre&ated su,Dpro'esses are so+eti+es re!erred to as a" IETLJ pro'ess. 21 E+tractD Si"'e data !or the data warehouse 'a" 'o+e !ro+ di!!ere"t sour'es a"d +a ,e o! di!!ere"t t pes) the p&a" to e*tra't the data a&o"# with appropriate 'o+pressio" a"d e"'r ptio" te'h"i.ues is a" i+porta"t re.uire+e"t !or 'o"sideratio". F1 Transform2 Tra"s!or+atio" o! data with appropriate 'o"(ersio") a##re#atio" a"d '&ea"i"# ,esides deD"or+a&i$atio" a"d surro#ate /e +a"a#e+e"t is a&so a" i+porta"t pro'ess to ,e p&a""ed !or ,ui&di"# a data warehouse. ;1 3oad2 Steps to ,e 'o"sidered to &oad data with opti+i$atio" , 'o"sideri"# the +u&tip&e areas where the data is tar#eted to ,e &oaded a"d retrie(ed is a&so a" i+porta"t part o! the data warehouse ar'hite'ture p&a".

Data 4odel Architecture

I" Data Mode& Ar'hite'ture 0a&so /"ow" as Di+e"sio"a& Data Mode&1) there are ; +ai" data +ode&i"# st &es !or e"terprise warehouses:D

5rd ,ormal 6orm DTop Dow" Ar'hite'ture) Top Dow" I+p&e+e"tatio" 6ederated Star Schemas D Botto+ Up Ar'hite'ture) Botto+ Up I+p&e+e"tatio" Data .ault D Top Dow" Ar'hite'ture) Botto+ Up I+p&e+e"tatio"

Technolog) Architecture
Te'h"o&o# or Te'h"i'a& ar'hite'ture pri+ar e(o&(ed !ro+ deri(atio"s !ro+ the pro'ess ar'hite'ture) Meta data +a"a#e+e"t re.uire+e"ts ,ased o" ,usi"ess ru&es a"d se'urit &e(e&s i+p&e+e"tatio"s a"d te'h"o&o# too& spe'i!i' e(a&uatio". Besides these) the Te'h"o&o# ar'hite'ture a&so &oo/s i"to the (arious te'h"o&o# i+p&e+e"tatio" sta"dards i" data,ase +a"a#e+e"t) data,ase 'o""e'ti(it proto'o&s 0ODB8) @DB8) OLE DB et'1) Midd&eware 0,ased

o" O>B) >MI) 8OMHDOM et'.1) Networ/ proto'o&s 0DNS) LDAP et'1 a"d other re&ated te'h"o&o#ies.

Information Architecture
I"!or+atio" Ar'hite'ture is the pro'ess o! tra"s&ati"# the i"!or+atio" !ro+ o"e !or+ to a"other i" a step , step se.ue"'e so as to +a"a#e the stora#e) retrie(a&) +odi!i'atio" a"d de&etio" o! the data i" the data warehouse.

Resource Architecture
>esour'e ar'hite'ture is re&ated to so!tware ar'hite'ture i" that +a" resour'es 'o+e !ro+ so!tware resour'es. >esour'es are i+porta"t ,e'ause the he&p deter+i"e per!or+a"'e. Bor/&oad is the other part o! the e.uatio". I! ou ha(e e"ou#h resour'es to 'o+p&ete the wor/&oad i" the ri#ht a+ou"t o! ti+e) the" per!or+a"'e wi&& ,e hi#h. I! there are "ot e"ou#h resour'es !or the wor/&oad) the" per!or+a"'e wi&& ,e &ow.

DATA7ASE
A data,ase is a" app&i'atio" that +a"a#es data a"d a&&ows !ast stora#e a"d retrie(a& o! that data.
The term database was originally written as data base, and it may have been first used in 1963 at a symposium sponsored by the System Development Corporation of Santa oni!a, California. The use of the term database "single word# be!ame popular in some $uropean !ountries in the early 19%&s, and it subse'uently spread to the (.S.

8A data ase is a collection of information that is organi9ed so that it can easil) e accessed, managed, and u$dated& In one view, data ases can e classified according to t)$es of content* i liogra$hic, full2te+t, numeric, and images&' A data,ase 'a" #e"era&& ,e &oo/ed at as ,ei"# a 'o&&e'tio" o! re'ords) ea'h o! whi'h 'o"tai"s o"e or +ore !ie&ds 0i.e.) pie'es o! data1 a,out so+e e"tit 0i.e.) o,9e't1) su'h as a perso") or#a"i$atio") 'it ) produ't) wor/ o! art) re'ipe) 'he+i'a&) or se.ue"'e o! DNA. Kor e*a+p&e) the !ie&ds !or a data,ase that is a,out peop&e who wor/ !or a spe'i!i' 'o+pa" +i#ht i"'&ude the "a+e) e+p&o ee ide"ti!i'atio" "u+,er) address) te&epho"e "u+,er) date e+p&o +e"t started) positio" a"d sa&ar !or ea'h wor/er.

T:1ES
There are di!!ere"t t pes o! data,ase ,ut the +ost popu&ar is a re&atio"a& data,ase that stores data i" ta,&es where ea'h row i" the ta,&e ho&ds the sa+e sort o! i"!or+atio".

Relational Data ase

The sta"dard o! ,usi"ess 'o+puti"# as o! F553) re&atio"a& data,ases are the +ost 'o++o"& used data,ase toda . It uses the ta,&e to stru'ture i"!or+atio" so that it 'a" ,e readi& a"d easi& sear'hed throu#h. A re&atio"a& data,ase is a wa o! or#a"i$i"# data su'h that it appears to the user to ,e stored i" a series o! i"terre&ated ta,&es. I"terest i" this +ode& was i"itia&& 'o"!i"ed to a'ade+ia) perhaps ,e'ause the theoreti'a& ,asis is "ot eas to u"dersta"d) a"d thus the !irst 'o++er'ia& produ'ts) Ora'&e a"d DBF) did "ot appear u"ti& arou"d 23?5. Su,se.ue"t& ) re&atio"a& data,ases ,e'a+e the do+i"a"t t pe !or hi#h per!or+a"'e

app&i'atio"s ,e'ause o! their e!!i'ie"' ) ease o! use) a"d a,i&it to per!or+ a (ariet o! use!u& tas/s that had "ot ,ee" ori#i"a&& e"(isio"ed.

O$erational data ase

These data,ases store detai&ed data "eeded to support the operatio"s o! a" e"tire or#a"i$atio". The are a&so 'a&&ed su,9e'tDarea data,ases 0SADB1) tra"sa'tio" data,ases) a"d produ'tio" data,ases. Kor e*a+p&e:

'usto+er data,ase perso"a& data,ase i"(e"tor data,ase

a''ou"ti"# data,ase

Anal)tical data ase

A"a& ti' data,ases 0a./.a. OLAPD O" Li"e A"a& ti'a& Pro'essi"#1 are pri+ari& stati') readD o"& data,ases whi'h store ar'hi(ed) histori'a& data used !or a"a& sis. Kor e*a+p&e) a 'o+pa" +i#ht store sa&es re'ords o(er the &ast te" ears i" a" a"a& ti' data,ase a"d use that data,ase to a"a& $e +ar/eti"# strate#ies i" re&atio"ship to de+o#raphi's. O" the we,) ou wi&& o!te" see a"a& ti' data,ases i" the !or+ o! i"(e"tor 'ata&o#s su'h as the o"e show" pre(ious& !ro+ A+a$o".'o+. A" i"(e"tor 'ata&o# a"a& ti'a& data,ase usua&& ho&ds des'ripti(e i"!or+atio" a,out a&& a(ai&a,&e produ'ts i" the i"(e"tor . Be, pa#es are #e"erated d "a+i'a&& , .uer i"# the &ist o! a(ai&a,&e produ'ts i" the i"(e"tor a#ai"st so+e sear'h para+eters. The d "a+i'a&& D#e"erated pa#e wi&& disp&a the i"!or+atio" a,out ea'h ite+ 0su'h as tit&e) author) ISBN1 whi'h is stored i" the data,ase.

Data warehouse
A data warehouse stores data !ro+ 'urre"t a"d pre(ious ears 6 data e*tra'ted !ro+ the (arious operatio"a& data,ases o! a" or#a"i$atio". It ,e'o+es the 'e"tra& sour'e o! data that has ,ee" s'ree"ed) edited) sta"dardi$ed a"d i"te#rated so that it 'a" ,e used , +a"a#ers a"d other e"dDuser pro!essio"a&s throu#hout a" or#a"i$atio". Data warehouses are 'hara'teri$ed , ,ei"# s&ow to i"sert i"to ,ut !ast to retrie(e !ro+. >e'e"t de(e&op+e"ts i" data warehousi"# ha(e &ed to the use o! a Shared "othi"# ar'hite'ture to !a'i&itate e*tre+e s'a&i"#.

Distri uted data ase

These are data,ases o! &o'a& wor/D#roups a"d depart+e"ts at re#io"a& o!!i'es) ,ra"'h o!!i'es) +a"u!a'turi"# p&a"ts a"d other wor/ sites. These data,ases 'a" i"'&ude se#+e"ts o! ,oth

'o++o" operatio"a& a"d 'o++o" user data,ases) as we&& as data #e"erated a"d used o"& at a user-s ow" site.

End2user data ase

These data,ases 'o"sist o! a (ariet o! data !i&es de(e&oped , e"dDusers at their wor/statio"s. E*a+p&es o! these are 'o&&e'tio"s o! do'u+e"ts i" spreadsheets) word pro'essi"# a"d e(e" dow"&oaded !i&es.

E+ternal data ase

These data,ases pro(ide a''ess to e*ter"a&) pri(ate& Dow"ed data o"&i"e 6 a(ai&a,&e !or a !ee to e"dDusers a"d or#a"i$atio"s !ro+ 'o++er'ia& ser(i'es. A''ess to a wea&th o! i"!or+atio" !ro+ e*ter"a& data,ase is a(ai&a,&e !or a !ee !ro+ 'o++er'ia& o"&i"e ser(i'es a"d with or without 'har#e !ro+ +a" sour'es i" the I"ter"et.

H)$ermedia data ases on the we

These are a set o! i"ter'o""e'ted +u&ti+edia pa#es at a we,Dsite. The 'o"sist o! a ho+e pa#e a"d other h per&i"/ed pa#es o! +u&ti+edia or +i*ed +edia su'h as te*t) #raphi') photo#raphi' i+a#es) (ideo '&ips) audio et'.

,avigational data ase

I" "a(i#atio"a& data,ases) .ueries !i"d o,9e'ts pri+ari& , !o&&owi"# re!ere"'es !ro+ other o,9e'ts. Traditio"a&& "a(i#atio"a& i"ter!a'es are pro'edura&) thou#h o"e 'ou&d 'hara'teri$e so+e +oder" s ste+s &i/e L Path as ,ei"# si+u&ta"eous& "a(i#atio"a& a"d de'&arati(e.

In2memor) data ases

I"D+e+or data,ases pri+ari& re& o" +ai" +e+or !or 'o+puter data stora#e. This

'o"trasts with data,ase +a"a#e+e"t s ste+s whi'h e+p&o a dis/D,ased stora#e +e'ha"is+. Mai" +e+or data,ases are !aster tha" dis/Dopti+i$ed data,ases si"'e the i"ter"a& opti+i$atio" a&#orith+s are si+p&er a"d e*e'ute !ewer 8PU i"stru'tio"s. A''essi"# data i" +e+or pro(ides !aster a"d +ore predi'ta,&e per!or+a"'e tha" dis/. I" app&i'atio"s where

respo"se ti+e is 'riti'a&) su'h as te&e'o++u"i'atio"s "etwor/ e.uip+e"t that operates e+er#e"' s ste+s) +ai" +e+or data,ases are o!te" used.

Document2oriented data ase

Do'u+e"tDorie"ted data,ases are 'o+puter pro#ra+s desi#"ed !or do'u+e"tDorie"ted app&i'atio"s. These s ste+s +a ,e i+p&e+e"ted as a &a er a,o(e a re&atio"a& data,ase or a" o,9e't data,ase. As opposed to re&atio"a& data,ases) do'u+e"tD,ased data,ases do "ot store data i" ta,&es with u"i!or+ si$ed !ie&ds !or ea'h re'ord. I"stead) the store ea'h re'ord as a do'u+e"t that has 'ertai" 'hara'teristi's. A" "u+,er o! !ie&ds o! a" &e"#th 'a" ,e added to a do'u+e"t. Kie&ds 'a" a&so 'o"tai" +u&tip&e pie'es o! data.

Real2time data ases

A rea&Dti+e data,ase is a pro'essi"# s ste+ desi#"ed to ha"d&e wor/&oads whose state +a 'ha"#e 'o"sta"t& . This di!!ers !ro+ traditio"a& data,ases 'o"tai"i"# persiste"t data) +ost& u"a!!e'ted , ti+e. Kor e*a+p&e) a sto'/ +ar/et 'ha"#es rapid& a"d d "a+i'a&& . >ea&Dti+e pro'essi"# +ea"s that a tra"sa'tio" is pro'essed !ast e"ou#h !or the resu&t to 'o+e ,a'/ a"d ,e a'ted o" ri#ht awa . >ea&Dti+e data,ases are use!u& !or a''ou"ti"#) ,a"/i"#) &aw) +edi'a& re'ords) +u&tiD+edia) pro'ess 'o"tro&) reser(atio" s ste+s) a"d s'ie"ti!i' data a"a& sis. As 'o+puters i"'rease i" power a"d 'a" store +ore data) rea&Dti+e data,ases ,e'o+e i"te#rated i"to so'iet a"d are e+p&o ed i" +a" app&i'atio"s.

AR-HITE-TURE
A "u+,er o! data,ase ar'hite'tures e*ist. Ma" data,ases use a 'o+,i"atio" o! strate#ies. Data,ases 'o"sist o! so!twareD,ased A'o"tai"ersA that are stru'tured to 'o&&e't a"d store i"!or+atio" so users 'a" retrie(e) add) update or re+o(e su'h i"!or+atio" i" a" auto+ati' !ashio". Data,ase pro#ra+s are desi#"ed !or users so that the 'a" add or de&ete a" i"!or+atio" "eeded. The stru'ture o! a data,ase is ta,u&ar) 'o"sisti"# o! rows a"d 'o&u+"s o! i"!or+atio". O"&i"e Tra"sa'tio" Pro'essi"# s ste+s 0OLTP1 o!te" use a Arow orie"tedA or a" Ao,9e't orie"tedA data store ar'hite'ture) whereas dataDwarehouse a"d other retrie(a& !o'used app&i'atio"s &i/e 7oo#&e%s Bi#Ta,&e) or ,i,&io#raphi' data,ase 0&i,rar 'ata&o#1 s ste+s +a use a 8o&u+" orie"ted DBMS ar'hite'ture.

Do'u+e"tDOrie"ted) LML) /"ow&ed#e,ase) as we&& as !ra+e data,ases a"d >DKDstores 0a&so /"ow" as trip&e stores1) +a a&so use a 'o+,i"atio" o! these ar'hite'tures i" their i+p&e+e"tatio" Not a&& data,ases ha(e or "eed a data,ase s'he+a 0As'he+aD&ess data,asesA1. O(er +a" ears #e"era&Dpurpose data,ase s ste+s ha(e do+i"ated the data,ase i"dustr . These o!!er a wide ra"#e o! !u"'tio"s) app&i'a,&e to +a" ) i! "ot +ost 'ir'u+sta"'es i" +oder" data pro'essi"#. These ha(e ,ee" e"ha"'ed with e*te"si,&e data t pes 0pio"eered i" the Post#reSML pro9e't1 to a&&ow de(e&op+e"t o! a (er wide ra"#e o! app&i'atio"s.

. There are a&so other t pes o! data,ases whi'h 'a""ot ,e '&assi!ied as re&atio"a& data,ases. Most "ota,&e is the o,9e't data,ase +a"a#e+e"t s ste+) whi'h stores &a"#ua#e o,9e'ts "ati(e& without usi"# a separate data de!i"itio" &a"#ua#e a"d without tra"s&ati"# i"to a separate stora#e s'he+a. U"&i/e re&atio"a& s ste+s) these o,9e't data,ases store the

re&atio"ship ,etwee" 'o+p&e* data t pes as part o! their stora#e +ode& i" a wa that does "ot re.uire ru"ti+e 'a&'u&atio" o! re&ated data usi"# re&atio"a& a&#e,ra e*e'utio" a&#orith+s.

Data ase management s)stems

Data,ase +a"a#e+e"t s ste+ 0DBMS1 'o"sists o! so!tware that or#a"i$es the stora#e o! data. A DBMS 'o"tro&s the 'reatio") +ai"te"a"'e) a"d use o! the data,ase stora#e stru'tures o! so'ia& or#a"i$atio"s a"d o! their users. It a&&ows or#a"i$atio"s to p&a'e 'o"tro& o! or#a"i$atio" wide data,ase de(e&op+e"t i" the ha"ds o! Data,ase Ad+i"istrators 0DBAs1 a"d other spe'ia&ists. I" &ar#e s ste+s) a DBMS a&&ows users a"d other so!tware to store a"d retrie(e data i" a stru'tured wa . Data,ase +a"a#e+e"t s ste+s are usua&& 'ate#ori$ed a''ordi"# to the data,ase +ode& that the support) su'h as the "etwor/) re&atio"a& or o,9e't +ode&. The +ode& te"ds to deter+i"e the .uer &a"#ua#es that are a(ai&a,&e to a''ess the data,ase. O"e 'o++o"& used .uer &a"#ua#e !or the re&atio"a& data,ase is SML) a&thou#h SML s "ta* a"d !u"'tio" 'a" (ar !ro+ o"e DBMS to a"other. A 'o++o" .uer &a"#ua#e !or the o,9e't data,ase is OML) a&thou#h "ot a&& (e"dors o! o,9e't data,ases i+p&e+e"t this) +a9orit o! the+ do i+p&e+e"t this +ethod. A #reat dea& o! the i"ter"a& e"#i"eeri"# o! a DBMS is i"depe"de"t o! the data +ode&) a"d is 'o"'er"ed with +a"a#i"# !a'tors su'h as per!or+a"'e) 'o"'urre"' ) i"te#rit ) a"d re'o(er !ro+ hardware !ai&ures. I" these areas there are &ar#e di!!ere"'es ,etwee" the produ'ts. A re&atio"a& data,ase +a"a#e+e"t s ste+ 0>DBMS1 i+p&e+e"ts !eatures o! the re&atio"a& +ode&. I" this 'o"te*t) Date%s AI"!or+atio" Pri"'ip&eA states: Athe e"tire i"!or+atio" 'o"te"t o! the data,ase is represe"ted i" o"e a"d o"& o"e wa . Na+e& as e*p&i'it (a&ues i" 'o&u+" positio"s 0attri,utes1 a"d rows i" re&atio"s 0tup&es1. There!ore) there are "o e*p&i'it poi"ters ,etwee" re&ated ta,&es.A This 'o"trasts with the o,9e't data,ase +a"a#e+e"t s ste+ 0ODBMS1) whi'h does store e*p&i'it poi"ters ,etwee" re&ated t pes.

-om$onents of D74S
A''ordi"# to the wi/i,oo/s ope"D'o"te"t te*t,oo/s) ADesi#" o! Mai" Me+or Data,ase S ste+HO(er(iew o! DBMSA) +ost DBMS as o! F553 i+p&e+e"t a re&atio"a& +ode&. Other &essDused DBMS s ste+s) su'h as the o,9e't DBMS) #e"era&& operate i" areas o! app&i'atio"D spe'i!i' data +a"a#e+e"t where per!or+a"'e a"d s'a&a,i&it ta/e hi#her priorit tha" the !&e*i,i&it o! ad hoc .uer 'apa,i&ities pro(ided (ia the re&atio"a&Da&#e,ra e*e'utio" a&#orith+s o! a re&atio"a& DBMS.

RDBMS components

Interface drivers D A user or app&i'atio" pro#ra+ i"itiates either s'he+a +odi!i'atio" or 'o"te"t +odi!i'atio". These dri(ers are ,ui&t o" top o! SML. The pro(ide +ethods to prepare state+e"ts e*e'ute state+e"ts) !et'h resu&ts) et'. E*a+p&es i"'&ude DDL) D8L) DML) ODB8) a"d @DB8. So+e (e"dors pro(ide &a"#ua#eDspe'i!i' proprietar i"ter!a'es. Kor e*a+p&e M SML a"d KireBird pro(ide dri(ers !or PNP) P tho") et'. S;3 engine D This 'o+po"e"t i"terprets a"d e*e'utes the SML .uer . It 'o+prises three +a9or 'o+po"e"ts 0'o+pi&er) opti+i$er) a"d e*e'utio" e"#i"e1. Transaction engine D Tra"sa'tio"s are se.ue"'es o! operatio"s that read or write data,ase e&e+e"ts) whi'h are #rouped to#ether. Relational engine D >e&atio"a& o,9e'ts su'h as Ta,&e) I"de*) a"d >e!ere"tia& i"te#rit 'o"strai"ts are i+p&e+e"ted i" this 'o+po"e"t. Storage engine D This 'o+po"e"t stores a"d retrie(es data re'ords. It a&so pro(ides a +e'ha"is+ to store +etadata a"d 'o"tro& i"!or+atio" su'h as u"do &o#s) redo &o#s) &o'/ ta,&es) et'.

ODBMS components

3anguage drivers D A user or app&i'atio" pro#ra+ i"itiates either s'he+a +odi!i'atio" or 'o"te"t +odi!i'atio" (ia the 'hose" pro#ra++i"# &a"#ua#e. The dri(ers the" pro(ide the +e'ha"is+ to +a"a#e o,9e't &i!e' '&e 'oup&i"# o! the app&i'atio" +e+or spa'e with the u"der& i"# persiste"t stora#e. E*a+p&es i"'&ude 8OO) @a(a) .NET) a"d >u, . ;uer) engine D This 'o+po"e"t i"terprets a"d e*e'utes &a"#ua#eDspe'i!i' .uer 'o++a"ds i" the !or+ o! OML) LINM) @DOML) @PAML) others. The .uer e"#i"e retur"s &a"#ua#e spe'i!i' 'o&&e'tio"s o! o,9e'ts whi'h satis! a .uer predi'ate e*pressed as &o#i'a& operators e.#. P) Q) PR) QR) AND) O>) NOT) 7roupBY) et'. Transaction engine D Tra"sa'tio"s are se.ue"'es o! operatio"s that read or write data,ase e&e+e"ts) whi'h are #rouped to#ether. The tra"sa'tio" e"#i"e is 'o"'er"ed with su'h thi"#s as data iso&atio" a"d 'o"siste"' i" the dri(er 'a'he a"d data (o&u+es , 'oordi"ati"# with the stora#e e"#i"e. Storage engine D This 'o+po"e"t stores a"d retrie(es o,9e'ts i" a" ar,itrari& 'o+p&e* +ode&. It a&so pro(ides a +e'ha"is+ to +a"a#e a"d store +etadata a"d 'o"tro& i"!or+atio" su'h as u"do &o#s) redo &o#s) &o'/ #raphs)