Lawm ntau hype nyob Hadoop ntev. Cov hype no yog tuaj vim tias Hadoop no perceived ib lub cuab tam heev npaum loj cov ntaub ntawv ua. Tab sis lub sij hawm tuaj saib tej no, tsis tseeb. Nws yog lub sij hawm thaum lub hype maj mam xyuas thiab lag luam saib cov xa rau peev pib (ROI). Ib daim ntawv ntsuam xyuas, los Gartner, nkawd los pom tias ntau lub tuam txhab uas muag cov tsis npaj nqis peev pab rau Hadoop vim tias lawv tsis muaj cov txawj siv tau nws los yog nws tseem tswv ib tug user-unfriendly uas. Muaj lwm yam zoo li. Txawm li cas los, muaj ib cov neeg muaj kev bullish txog cov cov prospects ntawm Hadoop. Yog hais tias qhov teeb meem no rau nkawd ruam tag, yog vim hais tias tus cwj pwm ntawm Hadoop yog yuav mus rau theem ntawm disillusionment ntawm cov hype thiab li tej. Qhov no nws yog lub sij hawm uas thaum cov lag luam pib tuaj muaj tiag. Qhov no nws yog lub sij hawm thaum tuam txhab uas muag yuav objectively soj ntsuam thiab siv cov Hadoop.
Contrasting tswvyim rau me nyuam los tu Hadoop
Raws li tau teev los saum no, cov lag luam yuav muab faib ua ob pawg thaum nws los txog rau me nyuam los tu Hadoop: ib pawg comprises qhauj tsis kam, tos los yog circumspect thaum nws los txog rau Hadoop los tu thiab cov pab pawg neeg thib ob comprises qhauj uas ntseeg tias Hadoop yog yuav muab ROI zoo. Tus cwj pwm ntawm cov lus qub reflected nyob rau hauv cov kev sojntsuam Gartner. Muab cov hauv qab no yog qhov nws tshawb pom salient los ntawm daim ntawv ntsuam xyuas cov. Nco ntsoov tias qhov nws tshawb pom hauv daim ntawv ntsuam xyuas tau txiav txim rau ntuj, 2015. Li ntawd, qhov txheeb xyuas tau cov zoo nkauj tshiab. Cov kev sojntsuam phiaj tuaj dub me me thiab nruab nrab tsuas tuam txhab uas muag thiab C-theem executives.
- 54% ntawm cov neeg tsis txhob txwm npaj mus nqis peev pab rau Hadoop yav tom ntej.
- Cia li 18% ntawm qhov cov neeg tau kev pab them nqi los sib piv cov nqi rau kev Hadoop nyob hauv ob lub xyoos tom ntej no.
- 26% ntawm tus neeg muaj deploying los xyaum nrog Hadoop.
- Tuam txhab uas muag uas tsis txaus siab ua los yog tos nrog Hadoop cov chaw tu txawj muaj tsis txaus thiab user-unfriendliness yog vim li cas tsis xav txog Hadoop.
Raws li Merv Adrian (Vice president nyob Gartner), “Muaj tej xwm txheej loj ntawm cov koom haum uas tsis muaj kev pab them nqi los sis twb txog lawv cov lus Hadoop, hnub qab nram ntsis demand kom zoo haum anemic nyob Hadoop tsawg ke 24 lub hli”. Tshaj tej ntawm, tus tsis muaj near-term kev npaj rau cov me nyuam los tu Hadoop uas qhia, dua li ntawm ntxiv tej yam txaus siab rau cov ntawv loj phenomenon, thov kom Hadoop yeej tseem tsis accelerating”. Vim li cas zoo li no tsis zoo lo lus teb mus rau Hadoop lub ntsiab yog muab hauv qab.
- Vim tsis txawj Hadoop yog ib qho tseem ceeb constraint. Qhauj thov tias lawv cov neeg ua haujlwm tsi muaj cuab kav siv Hadoop. Hadoop, nyob rau hauv nws daim ntawv thawj, muaj tau lom zem ntau uas ceev rau ib pawg twg kom leej twg yuav siv nws fab taws ntsos. Tab sis yog ib tug xov tooj ntawm lwm-tus neeg cov cuab yeej no los txog tsom txoj kev siv cov Hadoop, Txawm tias lawv yuav tsis yooj yim mus use. Tus tsis txaus siab loj tus Hadoop thiab cov cuab yeej lwm-tus neeg uas lawv yuav tsum tau cov kev txawj ntse tshiab kom tau learnt uas txhais tau tias ntxiv peev. Tsis tau yuav siv cov kev txawj ntse uas twb muaj lawm. Tab yog cov cuab yeej siv rau kev kawm, ntseeg tau tias nws yuav siv lwm 2 mus 3 xyoo rau cov kev pab cuam rau tsab credence.
- Rau ntau tus qhauj, Hadoop tsis yog qhov muaj feem thib ib. Lawv xav tias Hadoop yog overkill rau cov teeb meem ua hauj lwm nws yog thiab yuav kom daws tau. Nws zoo li deploying ib missile tua yoov ib pawg. Tseem, tus nqi rau cov me nyuam los tu Hadoop yog ntau tshaj rau cov kev pab derived ntawm dlaws lub lag luam teeb meem qhauj txojkev.
Cov pab pawg neeg thib ob yog optimistic thiab ntseeg txog tus me nyuam los tu Hadoop. Muaj cov tuam txhab uas muag uas muaj pib siv Hadoop rau lawv mainstream lag luam thiab reaping cov kev pab cuam. Lub ntsiab feature uas yog siv yog real-time ua cov ntaub ntawv. Piv txwv, tuam txhab uas muag yuav tiv thaiv kev dag ntxias ntawm cov ntsiab lus los ntawm ib tug real-time. Tuam txhab uas muag cov tau muab cov khoom zoo los cais rau ib real-time khaws cov ntaub ntawv tawm tswv yim lawv cov neeg muas zaub. Lawv yeej tau ntaub ntawv ntawm lub website-pab, yees duab, Internet tuamtxhab, kev tawm thiab ntau lwm qhov chaw. Nyob rau hauv ib daim ntawv ntsuam xyuas los TechValidate, 96% ntawm cov neeg uas khiav ob peb siv feem ntawm ib Hadoop ib pawg thiab co, 20% deploying ze li ntawm 50 siv neeg sawv Hadoop nkaus xwb. Cov kev sojntsuam ices tias 73% ntawm tus neeg muaj deploying lawv cov tshuaj thiab cov kev pab thiab 59% benefiting los ntawm kev txo nqi. Feem ntau lub tuam txhab saum toj no uas muag yog MapR cov neeg muas zaub. Raws li Bryon Dover, ib tug engineer loj cov ntaub ntawv nrog lub Rubicon Project, "MapR rau kuv qhov kev cia siab rau kom txheej txheem 3 trillion muas ib lub hli rau 99.999% uptime."
Dabtsi ua tau kom paub txog qhov nws tshawb pom saum toj no?
Tus ob huab miv ntawm Hadoop los tu yuav tau feem ntau yog ruam tag, tiam sis lub ntsiab lus teb ntawm lub lag luam cycle, qib no nruab nrab yog theem lwm xwb: cov qhauj tawm hauv lub hype tsis nyob ntev thiab yuav mus rau qhov kev luj xyuas nyob ntev. Thaum muaj hype, txhua yam rau nkawd liab tseb tab sis thaum twg muaj kev ntsuam xyuas, lub disadvantages tau tuaj. Li ntawd, cov lag luam nrhiav kev tawm zoo dua siv Hadoop rau lawv tej teeb meem ua hauj lwm los sis seb, Yog cov yuav tsum tau Hadoop.
Hadoop xav tau xwb nws exclusivity tseeb vim hais tias qhov ntawd yog ib tug uas siv yooj yim, muaj rau cov neeg uas yog chaw me lossis. Yog tsis muaj tus zoo front-end uas ua kom yooj yim rau cov neeg mus ua ntaub ntawv thiab kev txheeb xyuas cov ntaub ntawv. Tseem, yuav tsum mus kawm Hadoop thiab uas yuav tsum tau peev ntxiv. Thib cuab yeej hais tias yuav ua Hadoop tau yooj yim mus use tsis raws nraim nyob rau lawv cov neeg hu tuaj. Li ntawd, modifications xeeb thiab nws tus kheej yuav siv sij hawm. Yeej, Hadoop xav tau los ua pov thawj hais tias nws yog ib qho yooj yim mus use.
Qhauj tau twb paub lawm tias yog ib lub zoo tshaj siv ntawm Hadoop kev yog thaum koj yuav khiav cov ntaub ntawv ntawm lub sij hawm. Uas yog qhov chaw uas cov rau ob pawg ntawm cov neeg muas zaub yog reaping cov kev pab cuam. Batch ua yog tsis uas koj yuav tsum tau focusing. Ntawm lub sijhawm ua yuav pab kom koj tsis txhob frauds thiab muab cov tshuaj customized thiab cov kev pab rau koj cov neeg muas zaub. Tsis Hadoop meant rau cov ntaub ntawv zoo li qub. Cov duab hauv qab no qhia tau hais tias pab real-time yog tus muab biggest Hadoop.
Hauv no lub ntsiab lus teb, Nws tseem ev los pav Apache txim uas lawm tau ua ib txoj hauj lwm stellar cais loj cov ntaub ntawv ntawm lub sij hawm. Nws muab rau koj ib unified thiab tej qhov hau uas yuav pab koj tswj loj loj poob lawm cov ntaub ntawv los ntawm ntau qhov chaw hauv lub hauv paus real-time. Nrog cov txim biggest kom zoo dua yog tias koj yuav mus sau ntawv nab hab sej, Java los yog Scala thiab muaj ntau tshaj 80 tswv has. Nws kuj txhawb SQL queries, cov kev kawm tshuab, streaming cov ntaub ntawv thiab zauv ntawm teeb ntaub ntawv, sib dua daim ntawv qhia thiab txo tej haujlwm. Nyob rau hauv ib nutshell, nws yuav muab pov thawj los yog ib daim ntawv thov ua zoo real-time.
Kev pab ntawm lub sij hawm ntawm Hadoop
Image1: Kev pab ntawm lub sij hawm ntawm Hadoop
Tej me nyuam los tu lub tshuab tshiab yuav siv sij hawm. Hype thiab cov me nyuam los tu muaj ntau yam. Tseem tau uas feem pua ntawm cov 57% ntawm tus neeg ntawm tus Gartner daim ntawv ntsuam xyuas uas yeej tsis txhob txwm npaj mus nqis peev pab rau Hadoop yuav ua tom qab ib zaug twg raws li Hadoop sau ntau lawm mainstream theem ntawm tuam txhab uas muag ntau thiab cov nyiaj tau uas qhia. Qhov no yog ib qho uas tshwj xeeb yog tom qab tshiab yees cov khoom uas pib ua Hadoop ntau usable. Lub SQL nyob Hadoop, Piv txwv, tej zaum yuav yog starting cheej nyob ntawm Hadoop ua puas siv tau rau lub zej zog wider.
Txoj kev
Tus indifference ntawm Hadoop tsis tau nws tus tuam unproductive. Xwb tias lag luam yog tseem txawv nrog rau nws txoj kev. Raws li MapR cov neeg muas zaub yuav corroborate, koj yuav tsum qhia hais tias yuav zoo siv Hadoop rau dlaws koj cov teeb meem ua hauj lwm li cas. Nws siv rau real-time ua cov ntaub ntawv hauv mainstream zus pom tau qhov uas yuav mus. Piv, muaj lwm cov kev pab muaj heev uas tseem zoo li undiscovered. tau mas, ntau ntau Hadoop thiab cov ecosystem xav hloov. Nws yuav tsum ua kom puas siv tau rau cov neeg uas xav siv nws. Qeeb los tu tus nqi uas Hadoop ua dab tsi rau hauv lub taub Tawvv xam tom qab 2 mus 3 xyoo.
