Chapt

er1:I
ntr
oduct
ion

Dat
abaseSyst s,7th Ed.
em Concept
©Si
lber
schat
z,Kor
thandSudarshan
Seewww.
db-book.
com f
orcondi
ti
onsonr e-
use
 Dat
abaseAppl
icat
ionsExampl
es

 Enter
pri
seInfor
mation
• Sales:cust
omer s,pr
oducts,pur
chases
• Accounti
ng:payment s,r
eceipt
s,asset
s
• HumanResour ces:Infor
mat i
onaboutempl
oyees,sal
ari
es,payr
oll
t
axes.
 Banking
• Cust omeri nf
ormati
on,accounts,l
oans,andbanki
ngtransact
i
ons.
• Cr editcardtr
ansacti
ons
 Finance:
• Sal esandpur chasesoffinanci
ali
nst
ruments(e.
g.,s
tocksandbonds;
• St ori
ngr eal
-t
imemar ketdata

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
4 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseAppl
icat
ionsExampl
es(
Cont
.)

 Uni
v er
sit
ies:regi
str
ati
on,grades
 Ai
rl
ines:reservat
i
ons,schedules
 Manufacturi
ng:managementofpr oduct
i
on,i
nvent
ory
,or
der
s,suppl
y
chai
n.
 Tel
ecommuni
cat
i
on:r
ecor
dsofcal
l
s,t
ext
s,anddat
ausage,gener
ati
ng
mont
hlybi
l
ls,mai
ntai
ningbal
ancesonpr
epai
dcal
l
ingcar
ds
 Web-basedservices
• Onli
neretail
ers:ordert
racking,cust
omizedrecommendati
ons
• Onli
neadv erti
sements
 Documentdatabases
 Navi
gati
onsystems:f ormai
ntainingthelocat
ionsofvar
iespl
acesof
i
nter
estal
ongwi
t
htheex
actr
out
esofr
oads,t
rai
nsy
stems,buses,et
c.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
5 ©Si
lber
schat
z,Kor
thandSudar
shan
 Whydoweneedaspeci
ali
zedDB

 Whynotbuildt
heDBdi r
ect
l
yontopoffil
esys
tems.
 UsetheOSfilesy
stem
 Writ
etheprogr
amsaccessi
ngt
hedatausingaregul
arPL
• C,
• Java,
• Python
……

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
6 ©Si
lber
schat
z,Kor
thandSudar
shan
 Ear
lyDat
abaseSyst
ems

Int
heearl
ydays,dat
abaseappl
i
cat
i
onswer
ebui
l
tdi
rect
l
yont
opoffil
e
sys
tems,whi
chleadsto:

 Dat
aredundancyandi
nconsi
st
ency
:dat
aiss
tor
ed i
nmul
t
ipl
efil
e
f
ormat
sresul
t
ingi
ndupl
i
cat
i
onofi
nfor
mat
i
oni
ndi
ffer
entfil
es
 Difficultyinaccessingdata
• Needt owr i
teanewpr ogr
am tocarryouteachnewt
ask
 Dat ai solati
on
• Mul ti
plefil
esandformats
 I
ntegr i
typr obl
ems
• I ntegri
tycons t
rai
nts(e.g.
,accountbal
ance>0)become
“
bur
ied”i
npr
ogr
am coder
athert
hanbei
ngs
tat
edexpl
i
cit
l
y
• Har
dtoaddnewcons
trai
ntsorchangeexi
st
ingones

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
7 ©Si
lber
schat
z,Kor
thandSudar
shan
 Ear
lyDat
abaseSyst
ems(
Cont
.)

 Atomi
ci
tyofupdat
es
• Fai
lur
esmayleavedat
abasei
nani
nconsi
st
ents
tat
ewi
t
hpar
ti
al
updat
escar
ri
edout
• Exampl
e:Tr
ansf
eroff
undsf
rom oneaccountt
oanot
hershoul
dei
t
her
compl
eteornothappenatal
l
 Concur
rentaccessbymulti
pleuser
s
• Concurr
entaccessneededforperf
ormance
• Uncontr
oll
edconcurr
entaccessescanleadt
oinconsi
st
enci
es
 Ex:Twopeopl
ereadi
ngabal
ance(
say100)andupdat
i
ngi
tby
wi
t
hdr
awi
ngmoney(
say50each)att
hesamet
i
me
 Prot
ect
i
onprobl
ems
• Hardt
oprovi
deuseraccesst
osome,butnotal
l
,dat
a

Moder
nDat
abasesyst
emsoffersol
uti
onst
oal
ltheabovepr
obl
ems

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
8 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
aModel
s
 Acoll
ectionoft oolsfordescr
ibi
ng
• Data
• Datar elat
ionships
• Datasemant i
cs
• Datacons t
raints
 Manydifferentdatamodel s
• Relationalmodel
• Enti
ty -
Relationshipdatamodel(mai
nl
yf ordat
abasedesign)
• Object -
baseddat amodels(Obj
ect
-or
ientedandObject
-r
elat
ional
)
• Semi -str
uctureddatamodel( XML)
• Otherol dermodel s:
 Net
wor
kmodel
 Hi
erar
chi
cal
model

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
11 ©Si
lber
schat
z,Kor
thandSudar
shan
 Rel
ati
onalModel

 Al
lthedat
aisstoredinvar
ioust
ables.
 Exampl
eoftabulardatai
ntherel
ati
onalmodel
TedCodd
Col
umns
Tur
ingAwar
d1981

Rows

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
12 ©Si
lber
schat
z,Kor
thandSudar
shan
 ASampl
eRel
ati
onalDat
abase

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
13 ©Si
lber
schat
z,Kor
thandSudar
shan
 Vi
ew ofDat
a

Anar
chi
t
ect
uref
oradat
abasesy
stem

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
15 ©Si
lber
schat
z,Kor
thandSudar
shan
 Schemas

calSchema–t
 Logi heoveral
ll
ogi
cals
tructur
eofthedat
abase
• Exampl
e:Thedat
abaseconsi
st
sofinformat
i
onaboutasetof
cust
omer
sandaccount
sinabankandt
her
elat
i
onshi
pbet
ween
them
 Anal
ogoust
otypei
nfor
mat
i
onofav
ari
abl
einapr
ogr
am
calschema–t
 Physi heov
eral
lphy
sicals
truct
ureoft
hedat
abase

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
16 ©Si
lber
schat
z,Kor
thandSudar
shan
 I
nst
ancesandSchemas

 Si
milart
otypesandv ar
iablesi
nprogr
ammi nglanguages
 Schema–i sli
kea typeinPL
 I
nstance–theactualcontentoft
hedatabaseatapar t
icul
ar
poi
nti
nti
me
• Anal
ogoust
othev
alueofav
ari
abl
einapr
ogr
ammi
ng
l
anguages

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
17 ©Si
lber
schat
z,Kor
thandSudar
shan
 Physi
calDat
aIndependence

 Theabi
l
it
ytomodi
f
ythephy
sicalschemawi
t
houtc
hangi
ngt
he
l
ogi
cal
schema
• Appl
i
cat
ionsdependont hel
ogi
calschema
• I
ngener
al,t
heinter
facesbet
weenthev ar
iousl
evel
sand
componentsshoul
dbewelldefinedsothatchangesi
n
somepartsdonotser
iousl
yinfluenceot
hers.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
18 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
aDefini
ti
onLanguage(
DDL)

 Speci
ficat
i
onnot
ati
onf
ordefini
ngt
hedat
abaseschema
Ex
ampl
e: cr
eat
etabl
ei ns
tr or(
uct
ID char(5)
,
name var
char(20)
,
dept_name varchar(20)
,
salar
y numeric(8,
2))
 DDLcompi
l
ergener
atesasetoft
abl
etempl
atest
hatar
est
oredi
na
dat
adi
cti
onar
y
 Datadi
cti
onarycontainsmet
adat
a(i
.
e.,dat
aaboutdat
a)
• Databaseschema
• Int
egri
tyconst
raints
 Pr
imar
ykey(
exampl
e--I
Duni
quel
yident
i
fiesi
nst
ruct
ors)
• Aut
hor
izat
ion
 Whocanaccesswhat

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
19 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
aMani
pul
ati
onLanguage(
DML)

 Languagef
oraccessi
ngandupdat
i
ngt
hedat
aor
gani
zedbyt
he
appr
opr
iat
edat
amodel
• DMLal
soknownasquer
ylanguage
 Twocl
assesofl
anguages
e–usedf
• Pur orpr
ovi
ngpr
oper
ti
esaboutcomput
ati
onalpower
andf
oropt
i
mizat
i
on
 Rel
ati
onal
Algebr
a
 Tupl
erel
ati
onal
cal
cul
us
 Domai
nrel
ati
onalcal
cul
us
• Commer
cial–usedi
ncommer
cial
sys
tems
 SQLi
sthemos
twi
del
yusedcommer
cial
language

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
20 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
aMani
pul
ati
onLanguage(
Cont
.)

 Ther
earebasical
l
ytwotypesofdat
a-mani
pulat
ionl
anguage
• Pr alDML-
ocedur -requi
reausertospeci
fywhatdat
aareneeded
andhowt
ogett
hosedat
a.
• Decl
arat
iveDML -
-requi
reausert
ospeci
fywhatdat
aar
eneeded
wi
t
houtspeci
f
yinghowt
ogett
hosedat
a.
 Decl
arat
i
veDMLsar
eusual
l
yeasi
ert
olear
nanduset
hanar
epr
ocedur
al
DMLs.
 Decl
arat
i
v eDMLsar
ealsoref
err
edt
oasnon-pr
oceduralDMLs
 Theport
ionofaDMLthati
nvol
vesi
nfor
mat
i
onretr
iev
al i
scal
l
edaquer
y
l
anguage.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
21 ©Si
lber
schat
z,Kor
thandSudar
shan
 SQLQuer
yLanguage

 SQLi
saquer
ylanguagewhi
ch i
snonpr
ocedur
al.Aquer
ytak
esas
i
nputsev
eral
tabl
es(
possi
bl
yonl
yone)andal
way
sret
urnsasi
ngl
e
t
abl
e.
 Exampl
etofindt
henameofal
li
nst
ruct
orsi
nComp.Sci
.dept
selectname
fr
om instr
uctor
wher edept_name='
Comp.Sci
.
'
 SQLisNOTaTuri
ngmachi
neequi
val
entl
anguage
• Whatar
ethei
mpl
i
cati
ons?

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
22 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseAccessf
rom Appl
icat
ionPr
ogr
am

 SQLdoesnotsuppor
tact
i
onssuchasi
nputf
rom user
s,out
putt
o
di
spl
ays,orcommuni
cat
i
onov
ert
henet
wor
k.
 Suchcomput
ati
onsandact
ionsmus
tbewr
it
teni
nahostl
anguage,
suchasC,C++,Java,orPyt
hon,wi
t
hembeddedSQLquer
iest
hat
accesst
hedat
ainthedatabase.
 Appl
icat
ionpr ams-
ogr -ar
epr
ogr
amst
hatar
eusedt
oint
eractwi
t
h
t
hedat
abasei
nthi
sfashi
on.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
23 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseDesi
gn

Thepr
ocessofdesi
gni
ngt
hegener
als
truct
ureoft
hedat
abase:

 Logi
calDesi
gn– Deci
dingont
hedatabaseschema.
 Physi
calDesi
gn–Deci
dingonthephysi
call
ayoutoft
hedat
abase

 Dat
abasedesi
gnr
equi
rest
hatwefinda“
good”col
l
ect
i
onofr
elat
i
on
schemas.
• Busi
nessdeci
si
on–Whatat
tr
ibut
esshoul
dwer
ecor
dint
he
dat
abase?
• Comput
erSci
encedeci
si
on– Whatr
elat
i
onschemasshoul
dwe
haveandhowshouldt
heat
tr
ibut
esbedi
st
ri
but
edamongt
hev
ari
ous
rel
ati
onschemas?

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
24 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseEngi
ne

 Adat
abasesy
stem i
spar
ti
ti
onedi
ntomodul
est
hatdeal
wit
h
eachoft
her
esponsi
bil
i
ti
esoft
heov
eral
lsy
stem.
 Thef
unct
i
onal
component
sofadat
abasesy
stem canbe
di
vi
dedi
nto
• Thest
oragemanager,
• Thequerypr
ocessorcomponent
,
• Thetr
ansact
i
onmanagementcomponent
.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
25 ©Si
lber
schat
z,Kor
thandSudar
shan
 St
orageManager

 Apr
ogr
am modul
ethatpr
ovi
dest
hei
nter
facebet
weent
hel
ow-
lev
eldat
a
st
oredinthedat
abaseandt
heappl
i
cat
i
onpr
ogr
amsandquer
ies
submit
tedtot
hesystem.
 Thestor agemanageri sresponsi
blet
othefol
l
owingtask
s:
• Int
er actionwiththeOSfilemanager
• Efficients t
ori
ng,ret
ri
evingandupdati
ngofdata
 Thestor agemanagercomponent si
ncl
ude:
• Aut horizati
onandi nt
egrit
ymanager
• Transact ionmanager
• Fil
emanager( dataondi sk)
• Buffermanager( datainmai nmemory)

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
26 ©Si
lber
schat
z,Kor
thandSudar
shan
 St
orageManager(
Cont
.)

 Thest
oragemanageri
mpl
ement
ssev
eral
dat
ast
ruct
uresaspar
toft
he
phy
sicalsy
stem i
mpl
ement
ati
on:
• Dat
afil
es--st
oret
hedatabasei
tsel
f
• Dat
adicti
onar
y--stor
esmetadataaboutt
hes
truc
tur
eoft
he
dat
abase,i
npar
ti
cul
art
heschemaoft
hedat
abase.
• I
ndi
ces-
-canpr
ovi
def
astaccesst
odat
ait
ems.Adat
abasei
ndex
pr
ovi
despoi
nter
stot
hosedat
ait
emst
hathol
dapar
ti
cul
arv
alue.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
27 ©Si
lber
schat
z,Kor
thandSudar
shan
 Quer
yPr
ocessorComponent
s

 DDLi
nter
pret
er-
-int
erpr
etsDDLs
tat
ement
sandr
ecor
dst
he
defini
t
ionsi
nthedat
adi
ct
ionar
y.
 DMLcompi
l
er-
-tr
ansl
atesDMLst
atement
sinaquer
ylanguagei
nto
aneval
uat
i
onpl
anconsi
stingofl
ow-
lev
eli
nst
ruct
i
onst
hatt
hequer
y
eval
uat
i
onengi
neunder
stands.
• TheDMLcompi
l
erper
for
msquer
yopt
i
mizat
ion;t
hati
s,i
tpi
cks
t
helowestcos
tev
aluat
i
onpl
anf
rom amongt
hev
ari
ous
al
ter
nat
iv
es.
 Quer
yev
aluat
ionengi
ne-
-ex
ecut
esl
ow-
lev
eli
nst
ruct
i
onsgener
ated
byt
heDMLcompi
l
er.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
28 ©Si
lber
schat
z,Kor
thandSudar
shan
 Quer
yPr
ocessi
ng

1.Par
singandt
ransl
ati
on
2.Opt
i
mizat
i
on
3.Ev
aluat
i
on

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
29 ©Si
lber
schat
z,Kor
thandSudar
shan
 Tr
ansact
ionManagement

 At
ransact
ioni
sacol
l
ect
i
onofoper
ati
onst
hatper
for
msasi
ngl
e
l
ogi
cal
funct
i
oni
nadat
abaseappl
i
cat
i
on
 Tr
ansact
i managementcomponentensur
on- est
hatt
hedat
abase
remai
nsinaconsist
ent(cor
rect
)s t
atedespi
tesys
tem f
ail
ures(
e.g.,
powerf
ail
uresandoperati
ngsystem cr
ashes)andtr
ansacti
onf
ailur
es.
 Concur
rency-
cont
rolmanagercont
rol
sthei
nter
act
i
onamongt
he
concur
rentt
ransact
ions,t
oensur
etheconsi
st
encyoft
hedat
abase.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
30 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseAr
chi
tect
ure

 Cent
ral
i
zeddat
abases
• Onetoafewcor
es,sharedmemory
 Cli
ent
-ser
ver
,
• Oneserv
ermachi
neex ecut
eswor
konbehal
fofmul
t
ipl
ecl
i
ent
machi
nes.
 Par
all
eldat
abases
• Manycoreshar edmemor y
• Shareddisk
• Sharednothing
 Dist
ri
but
eddat abases
• Geographicaldis
tri
but
ion
• Schema/dat aheter
ogenei
ty

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
31 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseAppl
icat
ions

Dat
abaseappl
i
cat
i
onsar
eusual
l
ypar
ti
ti
onedi
ntot
woort
hreepar
ts
 Two-
ti
erar
chi
t
ect
ure-
-theappl
i
cat
i
onr
esi
desatt
hecl
i
ent
machi
ne,wher
eiti
nvok
esdat
abasesy
stem f
unct
i
onal
i
tyatt
he
ser
vermachi
ne
 Thr
ee-
ti
erar
chi
t
ect
ure-
-thecl
i
entmachi
neact
sasaf
rontend
anddoesnotcont
ainanydi
rectdat
abasecal
l
s.
• Thecl
i
entendcommuni
cat
eswi
t
hanappl
i
cat
i
onser
ver
,
usual
l
ythr
oughaf
ormsi
nter
face.
• Theappl
i
cat
i
onser
veri
ntur
ncommuni
cat
eswi
t
hadat
abase
sy
stem t
oaccessdat
a.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
32 ©Si
lber
schat
z,Kor
thandSudar
shan
 Two-
ti
erandt
hree-
ti
erar
chi
tect
ures

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
33 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseUser
s

Ther
ear
efourdi
ffer
entt
ypesofdat
abase-
sys
tem user
s

 Nai
veuser
s--unsophi
st
icat
eduser
swhoi
nter
actwi
t
hthesy
stem by
i
nvokingoneoft
heappl
i
cat
i
onpr
ogr
amst
hathav
ebeenwr
it
ten
pr
eviousl
y.
 Appl
i
cat
ionpr
ogr
ammer
s--comput
erpr
ofessi
onal
swhowr
it
e
appl
i
cat
i
onpr
ogr
ams.
 Sophi
sti
cat
edusers--i
nter
actwit
ht hesy stem wi
thoutwri
ti
ngpr ogr
ams
• Usingadatabasequeryl
anguageorby
• Usingt
oolssuchasdataanalysissoftware.
 Speci
ali
zedusers--
wri
tespeci
ali
zeddat abaseappl i
cat
ionsthatdonot
fitint
othet
radi
ti
onaldata-
processi
ngf
ramewor
k.Forex
ampl
e,CAD,
gr aphi
cdat
a,audio,vi
deo.

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
34 ©Si
lber
schat
z,Kor
thandSudar
shan
 Dat
abaseAdmi
nist
rat
or

Aper
sonwhohascent
ral
cont
rol
overthesyst
em i
scal
l
edadat
abase
admi
nist
rat
or(
DBA),whosef
uncti
onsare:
 Schemadefini
t
ion
 St
orages
truct
ureandaccess-
met
hoddefini
ti
on
 Schemaandphy
sical
-or
gani
zat
i
onmodi
ficat
ion
 Gr
ant
i
ngofaut
hor
izat
ionf
ordat
aaccess
 Rout
i
nemai
ntenance
 Per
iodi
cal
l
ybacki
ngupt
hedat
abase
 Ensur
ingthatenoughfr
eedi
skspacei
savai
l
ablef
ornor
mal
oper
ati
ons,andupgradi
ngdi
skspaceasr
equir
ed
 Moni
t
ori
ngj
obsr
unni
ngont
hedat
abaseandensur
ingt
hat
per
for
mancei
snotdegr
adedbyv
eryexpensi
vet
ask
ssubmi
t
tedby
someuser
s

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
35 ©Si
lber
schat
z,Kor
thandSudar
shan
 Hi
stor
yofDat
abaseSyst
ems

 1950sandear
ly1960s:
• Datapr
ocessi
ngusingmagnet
i
ctapesf
ors
tor
age
 Tapespr
ovi
dedonl
ysequent
i
alaccess
• Punchedcar
dsf
ori
nput
 Lat
e1960sand1970s:
• Harddi
sksal
loweddi r
ectaccesstodat
a
• Networ
kandhi er
archi
caldat
amodel si
nwidespr
eaduse
• TedCodddefinestherel
ati
onaldatamodel
 Woul
dwi
ntheACM Tur
ingAwar
dfort
hiswor
k
 I
BM Resear
chbegi
nsSy
stem Rpr
otot
ype
 UCBer
kel
ey(
Michael
Stonebr
eak
er)begi
nsI
ngr
espr
otot
ype
 Or
acl
erel
easesfir
stcommer
cial
rel
ati
onaldat
abase
• Hi
gh-
per
for
mance(
fort
heer
a)t
ransact
i
onpr
ocessi
ng

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
36 ©Si
lber
schat
z,Kor
thandSudar
shan
 Hi
stor
yofDat
abaseSyst
ems(
Cont
.)

 1980s:
• Resear
chr
elat
i
onalpr
otot
ypesev
olv
eint
ocommer
cialsy
stems
 SQLbecomesi
ndus
tri
als
tandar
d
• Par
all
elanddi
st
ri
but
eddat
abasesy
stems
 Wi
sconsi
n,I
BM,Ter
adat
a
• Obj
ect
-or
ient
eddat
abasesy
stems
 1990s:
• Largedeci
si
onsuppor
tanddata-
mini
ngappl
i
cat
i
ons
• Largemult
i-
ter
abyt
edat
awarehouses
• EmergenceofWebcommerce

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
37 ©Si
lber
schat
z,Kor
thandSudar
shan
 Hi
stor
yofDat
abaseSyst
ems(
Cont
.)

 2000s
• Bigdat
ast
oragesy
stems
 Googl
eBi
gTabl
e,YahooPNut
s,Amazon,
 “
NoSQL
”sy
stems.
• Bi
gdat
aanal
ysi
s:bey
ondSQL
 Mapr
educeandf
ri
ends
 2010s
• SQLr
eloaded
 SQLf
rontendt
oMapReducesy
stems
 Massi
vel
ypar
all
eldat
abasesy
stems
 Mul
ti
-cor
emai
n-memor
ydat
abases

Dat
abaseSyst
em Concept
s-7th Edi
ti
on 1.
38 ©Si
lber
schat
z,Kor
thandSudar
shan
EndofChapt
er1

Dat
abaseSyst s,7th Ed.
em Concept
©Si
lber
schat
z,Kor
thandSudarshan
Seewww.
db-book.
com f
orcondi
ti
onsonr e-
use