Enqubweni yenguquko yedijithali yomnotho, isintu kufanele sakhe izikhungo eziningi zokucubungula idatha. Izikhungo zedatha ngokwazo nazo kufanele ziguqulwe: izindaba zokubekezelela amaphutha kanye nokusebenza kahle kwamandla manje zibaluleke kakhulu kunangaphambili. Izinsiza zisebenzisa amanani amakhulu kagesi, futhi ukwehluleka kwengqalasizinda ye-IT ephakathi kwazo kubiza kakhulu amabhizinisi. Ubuhlakani bokwenziwa kanye nobuchwepheshe bokufunda kwemishini busiza onjiniyela - eminyakeni yakamuva buye basetshenziswa kakhulu ukudala izikhungo zedatha ezithuthuke kakhulu. Le ndlela ikhulisa ukutholakala kwezinsiza, inciphisa inani lokwehluleka futhi inciphise izindleko zokusebenza.
Isebenza kanjani?
Ubuhlakani bokwenziwa kanye nobuchwepheshe bokufunda komshini busetshenziselwa ukwenza izinqumo zokusebenza ngokuzenzakalelayo ngokusekelwe kudatha eqoqwe kuzinzwa ezihlukahlukene. Njengomthetho, amathuluzi anjalo ahlanganiswe nezinhlelo zekilasi ze-DCIM (Data Center Infrastructure Management) futhi akuvumela ukuba ubikezele ukwenzeka kwezimo eziphuthumayo, kanye nokuthuthukisa ukusebenza kwemishini ye-IT, ingqalasizinda yobunjiniyela kanye nabasebenzi besevisi. Ngokuvamile, abakhiqizi banikeza izinsizakalo zamafu kubanikazi besikhungo sedatha abaqoqa futhi bacubungule idatha evela kumakhasimende amaningi. Amasistimu anjalo enza kube jikelele umuzwa wokusebenzisa izikhungo zedatha ezihlukene, futhi ngenxa yalokho asebenza kangcono kunemikhiqizo yasendaweni.
Ukuphathwa kwengqalasizinda ye-IT
I-HPE iphromotha insizakalo yokuhlaziya yokubikezela ngamafu ukuphatha ingqalasizinda ye-IT eyakhelwe ku-Nimble Storage kanye nezinhlelo zokugcina ze-HPE 3PAR StoreServ, amaseva we-HPE ProLiant DL/ML/BL, amasistimu ama-rack we-HPE Apollo kanye nesiteji se-HPE Synergy. I-InfoSight ihlaziya ukufundwa kwezinzwa ezifakwe kumishini, icubungula imicimbi engaphezu kwesigidi ngomzuzwana futhi ihlale izifunda ngokwakho. Isevisi ayitholi kuphela amaphutha, kodwa futhi ibikezela izinkinga ezingase zibe khona nengqalasizinda ye-IT (ukwehluleka kwezisetshenziswa, ukukhathala komthamo wokugcina, ukwehla kokusebenza kwemishini ebonakalayo, njll.) ngisho nangaphambi kokuba zenzeke. Ngokuhlaziya okubikezelwayo, isofthiwe ye-VoltDB isetshenziswa efwini, kusetshenziswa amamodeli okubikezela okuzenzakalelayo nezindlela ezingenzeka. Isixazululo esifanayo siyatholakala kumasistimu wesitoreji esiyingxube kusuka ku-Tegile Systems: isevisi yefu ye-IntelliCare Cloud Analytics iqapha impilo, ukusebenza nokusetshenziswa kwezinsiza zamadivayisi. Ubuhlakani bokwenziwa kanye nobuchwepheshe bokufunda komshini nakho kusetshenziswa i-Dell EMC kuzixazululo zayo zekhompuyutha ezisebenza kahle. Kunezibonelo eziningi ezifanayo; cishe bonke abakhiqizi abahamba phambili bemishini yekhompuyutha nezinhlelo zokugcina idatha manje sebelandela le ndlela.
Ukunikezwa kwamandla nokupholisa
Enye indawo yokusetshenziswa kwe-AI ezikhungweni zedatha ihlobene nokuphathwa kwengqalasizinda yobunjiniyela futhi, ngaphezu kwakho konke, ukupholisa, isabelo lapho ukusetshenziswa kwamandla okuphelele kwesikhungo kungadlula ama-30%. I-Google ibingomunye wabokuqala ukucabanga ngokupholisa okuhlakaniphile: ngo-2016, kanye ne-DeepMind, yathuthukiswa. zokuqapha izingxenye zesikhungo sedatha ngazinye, ezehlise izindleko zamandla zokupholisa umoya ngo-40%. Ekuqaleni, yanikeza abasebenzi izeluleko kuphela, kodwa yabuye yathuthukiswa futhi manje isingakwazi ukulawula ukupholisa kwamagumbi emishini ngokuzimela. Inethiwekhi ye-neural efakwe emafini icubungula idatha evela ezinkulungwaneni zezinzwa zasendlini nezangaphandle: yenza izinqumo ngokucabangela umthwalo kumaseva, izinga lokushisa, kanye nesivinini somoya ngaphandle namanye amapharamitha amaningi. Iziqondiso ezinikezwa isistimu yamafu zithunyelwa esikhungweni sedatha futhi lapho ziphinde zihlolwe ukuphepha amasistimu wendawo, kuyilapho abasebenzi bengakwazi ukuvala imodi ezenzakalelayo futhi baqale ukuphatha ukupholisa ngesandla. I-Nlyte Software kanye nethimba le-IBM Watson idaliwe , eqoqa idatha ngezinga lokushisa kanye nomswakama, ukusetshenziswa kwamandla nokulayisha kumishini ye-IT. Ikuvumela ukuthi uthuthukise ukusebenza kwezinhlelo ezingaphansi zobunjiniyela futhi ayidingi ukuxhumeka kungqalasizinda yamafu yomkhiqizi - uma kunesidingo, isisombululo singafakwa ngqo esikhungweni sedatha.
Izibonelo eziningi
Kunezixazululo eziningi ezintsha ezihlakaniphile zezikhungo zedatha emakethe futhi ezintsha zihlala zivela. I-Wave2Wave idale isistimu yokushintsha ikhebula ye-robotic fiber optic ukuze ihlele ngokuzenzakalelayo ukuxhumana okuphambanayo ezindaweni zokushintshana kwethrafikhi (Meet Me Rooms) ngaphakathi kwesikhungo sedatha. Uhlelo olwakhiwe yi-ROOT Data Center kanye ne-LitBit isebenzisa i-AI ukuqapha amasethi ejeneretha e-diesel, futhi i-Romonet idale isisombululo sesofthiwe sokuzifundela sokuthuthukisa ingqalasizinda. Izixazululo ezidalwe i-Vigilent zisebenzisa umshini wokufunda ukubikezela ukwehluleka nokuthuthukisa izimo zokushisa emagcekeni esikhungo sedatha. Ukwethulwa kobuhlakani bokwenziwa, ukufunda komshini kanye nobunye ubuchwepheshe obusha bokwenza izinqubo ezizenzakalelayo ezikhungweni zedatha kwaqala maduzane, kodwa namuhla lena ingenye yezindawo ezithembisayo kakhulu zokuthuthukiswa kwemboni. Izikhungo zedatha zanamuhla sezinkulu kakhulu futhi zaba yinkimbinkimbi ukuthi zingaphathwa ngempumelelo ngesandla.
Source: www.habr.com