NoSQL is the New Hadoop

Information about NoSQL is the New Hadoop

Published on August 11, 2014

Author: tatabss

Source: authorstream.com

Content

PowerPoint Presentation: NoSQL i s t he New H a doop T h e k ey c hal le n ge b usi nesses w orld o v er face t o d ay is m a nagi n g d a ta e x plosi o n. T h e t r a d i tio n al busi n ess c o n ce pts t h at w ere used to ma n age da ta h ave b eco m e o b sol e te now. T h e c ha n g i n g dynamics in th e tec h no l ogical la n d s c a pe has l e d to newer a n d mo re s o phist i c a ted t o ols t h at w ork on da t a at j e t s p eed t hese day s . I’ v e f o und th a t emerg i ng tec h nol o gies li k e t h e New H a doop f ram e w ork aim at b etter s oluti o n for b ig d a ta s y st ems. W hy r el at ion a l d a tabase is not r ele v a nt a ny m o re ? Rela tio n al d a t a b ase m a nageme n t s y st em or RDBMS in t h e tr a di t i o n al set up h ave b e e n th e o nly o p t i on u s e d b y organ i z a tio n s t o ma n age t h eir da ta b ases e ff e c tively. The rela t i o nal d at a b ase hel p s to org a ni z e d a ta in a str u c tu r ed m a nn er based o n rel a tional m o d el. Thou g h I th i n k k eepi n g d a ta in a str u cture form is good f or e n te r p rises, in case of h u ge vol u mes t his c an b ecome a big bu r d e n , leadi n g t o p rogressive d ecli n e in p e r f orm a n c e. T h e s ce n e w ill b e mo r e fre q u e n t, o n ce t he d at a b ecomes t o o big to ma n age. T his ma k es R D BMS an i na p p r o pria t e s c ala b le so l u ti o n f or big d a ta. G e n e r ic D ata P roc ess ing F r a me wo r k Si n ce rel a tional d a ta base c o u ld n o t sa t isf y t h e d eman ds o f d a ta, a n a lt ern a ti v e sol uti o n w as requi r e d . T his re s ult e d i n t he i n tr o du c t i on of d a ta p r ocessing so f t w are. I ’ve had m a n y q ueries f rom t hose n ew to d a t a b ase ma n age m e n t a b ou t w h a t is H a do o p? It is nothi n g b u t s o f t w are f ram e w ork t h a t e na b l es paral l el p roc essi n g of hu ge amo u n ts o f d a ta i n a large co mmod i ty har d w are c l u s ter. The e n tire p r ocessi n g is err o r f r ee a n d un s w erving. T h e so f t w are c an exe c u t e qu er i es a n d a lso r ead o p era t i o n s on h u ge d a ta se t s, w hi c h h ave t h e c a pa b il i ty o f s c ali n g to as PowerPoint Presentation: Pr ud e nt Use of D a taba s es big as p e ta b y te sizes. The so f t w are f ra m ework has an unrivalled p rice p e r f orm a n ce ra tio t h at is b rou g h t a b o u t b y t he f lexible a n a l ytic s f e a tu r e it exh i bi t s. S t r u c t u r e d , sem i -st r u c t u red, a nd unst r u c t u red d a ta c a n be a n al y z ed w i t h t h e sa m e fi x e d fram e w or k . P a r a llelism and i t s Us e s T h e m ain a d va ntage of H a doop is i t s a b il i ty t o rou te paral l el qu e r ies in th e form o f hu g e b ac k gro u nd b a t ches w i t hin t h e same server f a rm. T h is redu ces t h e e x p e nses of us i n g a n a d d i tio n al h a r d w are as w as th e case in t r a d i ti o nal da t a b ase s y s tems. A nd i n my o p i ni o n , t h e time a n d e ff o rt needed is greatly re d u ce d . T h e c o n ce pt f or this type o f f ram e w ork origin a t e d f rom s earch e n gi nes li k e Ya h o o ! a n d G o ogle , w hi c h u se massive i n ex p e n sive servers to re a d parall e l q u er i es, so sea r c h i n di c es a n d rel a ted da ta st r u c tu r es c an b e formed. B u t w h e n t he d a ta to b e a naly zed b e c a me alarm i n gly hu g e in s i z e, th e s y s tem c ould not k eep u p as th e s c ali n g n eeded lo t s of coo r di n a t i n g an d c ach i n g m e t h o d s to r e d u ce t h e a lign m e n t requi re d . New H e ights in Sc a l a bi li ty T h e i n t r oducti o n o f n e w H a doop tec h nol o gy l i k e YARN (Y e t A n o th e r Re s o u rce Neg o tia t o r) has b rou g h t n ew h ei g h ts to t h e s c ala b ili t y f a c t o r o f t he fi l e s y st em. T h is n ew a ddi t ion has e n ha n c e d th e distr i b u tion p roc ess i n g of t h e s y st em w i t h t h e su cc essful m a n age m e nt o f big da ta. The highlig h t of t his n ew technol o gy is c lea r assigni ng of respo n s i bili t i es to d i f f e r e n t c ompone n ts, thus ma k i n g i t a hi g hly d esirable s y stem th a t I ’ d r eadily re c o m mend. D a tabase for D e a l i ng w i th Hi g h D ata Vo l u m e I’d suggest t h e emerg en c e of new da t a b a s es th a t are a p p r o p ri a te f o r un s tr u c t ur e d da t a is vital f or d a ta ma n age m e n t. W h a t is NoSQL? It is a n ew ge ner a tion da t a ba s e ma n agem e n t s y st em th a t e n a b les easy ac c ess a n d u ti l i z a tion of poly s tr u c t u r e d d a ta i n large vol u mes. So m e o f t he k ey points it a d d resses a re: C ost effe c tive s c ala b le s o l u ti o ns Fle x i b le assessme n t of da ta str u c t u res, w hi c h do not c o n f orm to t he re l a ti o n a l s y st em lik e gra p hs a n d k e y- va l ue i n f orm a tion T h e d a t a b ase p e r f o r ms a h oriz ontal ty pe of s c a l i n g c alled sh a r d i n g i n w hi c h eac h serv e r has a se p arate d a t a b ase t h at i s p ar ti t i o n e d ph ys i c ally, so eac h has th e d a ta st o red in t he local dis k s in i t . T h e d rawb ack I’ v e e x peri e n c ed h e re is you c a n n o t do jo i ns, s c hema c ha n ges or t ra n sa c t i ons a n d you may also n e e d to c o m p romise th e ACI D (Atomic i t y, c o n si s ten c y, isola t i o n , a n d du r a bili t y) w hi c h resu l ts i n relax i n g o f th e c o n sis t e n cy f actor.

Related presentations