c'oscakon Digilal ata

Ungutud data (unorga nir A

Semistuctird data

shuctuyd data

wih ghurA ata

Cae
 &ne Date Andyg*
choracterishes o data
4-Compostion t deals with he stchee ol dol
2. Condition !deale , th gtat ol data data.
3 Contee t : de als oith where has data bee

Euouton o Biq Data

tite DBMs
Delintin tiqh toume
gh -veoah
Biq data is highvdume, Wahvaiehy
highvelouh , high-voicy
inn asse l that demand Cost -e ecte
east e<]ectrue , Innov atue tnnovatue tormg at
inyn pouing faocesina
foenhanced in saht
deis on making Cnhanch ed inight
harnnex & decesion mahtna

Challenges iualia ahan "laphre ovage Starh

"Analsis
-Traner violaion
 Why biydata ?
)
claraleiatis o biq
data
3) Vedohy
\cose the ate at data (KB,B, GB
hich deta been
data oriqin acd ).
S
Nalidih toom
(înacurray) generaed.
Vdaik how
G lbriabuby change
dong Hhe hdooqereous
4(tHouo
date will be Sorrce.
acurat 'mto
vald
data to. CSpilkeg)Ctention
BT

data ie envirnment
housed n Ba dota enuionmert
Cenral ner
data analyr d in 8ygtem)
Oine modé Boh online a4
Nertial aling
a deale uoith ine
+ honontal gcalin
nd
dota wsth
4 Datu ie faken to data unehaduzd
ooaeing unchm uniong are taken t
data
ERRK
PepostngP
Doshboura
CRM2
Data wachouge HcLAP
odhoc quening

This d PaYy app

bakat hallengs hal poewnt Basin ay lurm
Cop:tal'sing big data.
on

donaton 1 data de dagux bonn

nashhee oraareohon
Dala ien lence ctacing
Bugines knandedge Joom Dala

Telhnolgs Jmahemeks
eperbe experb
Date Sienist
Businese acu men
Techmoloqs Eper tise
Aunderstandlin af domain ood database knowledge
Busineig hakgy >ood NOSQA B-1
Rnoblem Soloinq Dasay Pyhon, ct
> Çommuniaton >) Open 2ouce toole os
Roesentzia >Dat m Hodop
Inquis thueness Uis uaisahon ools

-Madine doorninq
s&hon,
}Nahtcd dnguage
 Respoei biih -Dada Manament .Analyhe tehniques
-Business Analyots

»in-Meron Analyhes Tesminodqie

.in-database porocSsing
"Symmet multpraesor gakme Massivey oalled paoessing
shee ho ase NosQ
Log analysis
+Time baseddata
nehoorking tude
key- Malue pais Graph -based coumn -0nened

Doument -bage d

2 Open Sarce
hadoop
dishi buted oeork lr shorage an dd
inherent dat
porotchs
paotsing gkrage
Hedbaay why Hado op? tast

&alabiliy powr
 HADoo?
Relotioncd Dotabase Node Basod at
byslern shuctu.
Managrnct
Data Suitable tor sb hocd
data data.
Ponoessing O1TP
Anayical, Big Dota
choice data need consishnt Bia data pnoeseigPorauing
dabonghip
oPen Source iy toameoork
ti Digtobuted huyMassiue Shovaqe
Hgdop Coe tomponen (uwrtt oneead meng)
skoraqe corporents ’ Hado op Diahibutrd tle sygteml trDFc)
Rgrace ssina sngine ’ Map teduce Rooromming
mcp a edue

do all o the aboue

Daw a
3> S

d. sbucherd dot
he abowe
Communiaon bl
ding
b) Biq deta ors to dat gs Hhet ne atdea
petabyte m
) Both data E, Cost etteche to mine
date ho make bugine Sense out
q) c) Reat hme
lo)
e abyuc
 Jlod oop VeesioD
VARN’ Resou re mQngq Aay
Vet Another

Apa dhe Hodoop Tussys tem

3 Cate qone D Data Snges ion
Sq0op Ha doop
ttume do q aga eqahor
( Data Tooeasing
Mcq nedee
Sporko
Data Andigis
pg
Ave Cuy hana age)
Cauoy kangu
împ ala

Ambar Manaqement
Mebout Cmachine Asorni
 1adoop eplica
Replicotion tatr 2

4st epuca placd tn, Hhe Sarne dGBcmode

Hhe Barne rack

9nd
epa is plaed allleert vack

the
isplaed cnoher allerrt node but
rak he and eplia
HDP3 lorsan
yhadoop
RDD (Reient Distibuted Datag
worker Node
Spa lontext eugto Manees SecurOs
Magter Node Task Tagle|[asl
 mgMabi

objeet D¢mo
Aey maán args : Amay Lsking3)
Neoy my Mati! 1DinTintJ(9, 3 )
Ver myMatsz o Din Cintj (3, 3)
Nar mymati3 Dim it) ,3)
B+4)

borl int 320,3a ol 33+4

mytMohie3jtjsro yMatn'ai ki3L3J x mymatia ol53rU

3

3
whilel count d
 Bubble Box

prl3oto n-)

Juoap(ami), arli41)33
3

3
obyet
 u-2

Appaaton o AesQL dotabag

a ime bage data

NogQ
ndhoodi
dhaatohe
ng
4 Non elatanal
No ed database &ystem
agkem.
table sehema
multiple dounnts
and bansachion
Reluzed
-AiD poopesha
ype ol NosQ databage
*key vaue paiy , pair MongoDB
Bhemo les Cassand, Coueh DB
Adwantas ousing Nosa
1- Can easy sale up and don

i darqe volwme al data

J- Dogot rqaied poe - defined schem

3.
cheap stass
4: Reloyes the data congistency vegui omert.
g- Data can be rpliuatkd to muliple
uliple no des
and can be pariboned .
 8QL NosQh
rel attonal dg non -xlahonad dB, dhibtA
relohon cl mocel nodel- less
Pone- dohned scherna dgamic heng
able buge data buse
key-vaue pair, aluon based
doumer base
wental sralfng
Soing saling mackine comnouh
Useg 3Q

tAP
aryuoge)
paeferc d ackened or Jarge
utaget Data

Sugports complei ocanot ase good support

" 3taong Consigkn5 tongi sBeny
tq: MonyoDB, tag gandre,
Postgnes 44bag e

Aladoop
teahens a tadoo
") han dly au
ype o data. ay ghare noting ahibea
F plcate data oLtpoLAP
good ohon Cannot be por allq2ed
8mal
 key- adoantase ( 0o tangsl ahion Ihong
fotm)

J> Sualeble
3 ost

Vevston ol Hadop
trsqtom n vehog
Data shonty
oameoh IrDrs

take key value poin

-Anban
CPoosisipni ng taonginq and monibring kodetp aushd
Mahout Aie
(RD CDota to)
Data Data Poocesi n |Cworl
Colectoy) Map cdeLe
tume Caistibutc Heage
Choq Zo0
Date
table Souru)
colecb (Hadoop destaibuted le Sygkn ) nua)
 todoop
irporin dotz orn DBng
Eaporg datz faom tudosp fle ,m.
uge3 Connecbr - based vctevse

Qnd
ollowing you b ahed,b
conae rnport ard erport taske

Log agggabr (aggveg to lqs om di<tent

aem
nd chine and lcace them nn HDF.S
desiqned tor igr voume ingu~
Data Paoe3sina
1. Map t duce Allos distäbukd ard penallel
osing huge datasets. Mapghose an d
rdee phase n HDER

Sponk!’nstern n Seala Cn memo omp uhn4)

workoads done
done inin memoy
memnoy vother thon dsk.
d
penk brav
 Data Anelys
hagh- dewd ping darguage
lain Pig deatn ar ronglated Gnto
MapRete gobe
he dta HDES Prto Piy
he nnime enuonment.

"Ronmaszchon thodoop
qging anyaie
Hive ls ROBMs
tave rad shema enlored
J ie ig a3ed wsite sdhoma entond
0Rt Once
On
and rea à many bme
Lheuoud
3. not ead ond ont mm hm km
O1T, oLAP ohne as RDBnS upport D7
4 Hibe gtat da vlg RDEMP daynci data.
S eagfy Bcalable easiy cala ble
Paraid tonputing le en a coriguting
RDM
TB,PB

theeis latenay due to

batch oo(as ng Gmmediale
Hme
mspone
ghene quixD
ad wnte
faeter Copcd, b ste
fastr
Cormpa d lo bd
 porocessing fastr
dads slor Massio Distibuted "
applh (dwdo? Loork
"frame 3lw dource Upen "
Hoadoop aspecs Key
-neduce meg
Data &btentqnate
copis dota o
zcbn Repleahon
Corniyuhng tailux
chaleng
by
orotrcho
Sera4lDistobuteB
eulbathy
data fnheen
Scalabidh " poer Computina
shoroge -agh cost Lo Hadoo
Sests ACID
tainq Vetend iaialh Low
Hl H
scalirg Hoizontal
kos
ditouey dotoayte9,
 HDTSTaemons Srnol
blo cte
l. Piee
is sloe 4
m racks )
N anmenado
y docta narmupa ce
(macder nodo)

TSimage +4 glox
log delails’ t stoes he
napsho
Co q detaile.
a coledon Data Node

ame ode
Dato ode

Lo cal
DIcle

Seondan Namenode (Name node

Duta
Hode Duta uta No de
Node
 5.get Blodk loation
1-0pern Jome no
+IDF& cç ent

FS detaínput
8htano
ient nole
S. kad

Data node Datno do

WORm onby rcad ran)

Analomg ol k wte

HDTS uend3:ot JNamenod

f8 dataDt put Ncmenode
stam
SAk Puket
Paket
Ipatanode pat dodd
Datuote Dutonode
Shatcqygame hot bt
the
node
seond
as
temend at:
uerent
on a
ent.
i
pla plaed
Repua oepuca
ieplaed Some
he
node
dert
ack
node
as
ertt
poresepu
ca pipeine
A
Second