- Izinhlelo ezisatshalaliswayo zisabalalisa ukucubungula kanye nedatha kuzo zonke izindawo eziningi ezihlanganisiwe, zithuthukisa ukusebenza, ukubekezelela amaphutha, kanye nokukhula.
- Ukwakheka kwayo kungaba yi-client-server, i-peer-to-peer, i-service-oriented noma i-microservices, okuhlanganisa ukuhlukaniswa kwedatha kanye nokukopishwa.
- Ziyisisekelo sezinsizakalo zamafu, ezentengiselwano ze-e-commerce, ezokuxhumana, ezamabhange, i-Big Data, i-AI, kanye namanethiwekhi e-IoT emhlabeni jikelele.
- Ukukhetha uhlelo olufanele olusatshalaliswayo kudinga ukuhlaziya ivolumu yedatha, isidingo esiphezulu, isabelomali, izikhathi zokuphendula, kanye nesu lokukhula.
Izinhlelo ezisatshalaliswayo zikhona yonke indawo , yize zivame ukunganakwa: njalo uma usesha okuthile ku-Google, ukhokha ngekhadi, usakaza uchungechunge, noma udlala umdlalo oku-inthanethi, uthembele kulolu hlobo lokwakha ngaphandle kokuqaphela. Ziyisisekelo esithule somnotho wedijithali wanamuhla futhi zivumela izigidi zabasebenzisi ukufinyelela isevisi ngasikhathi sinye ngaphandle kokuthi lonke uhlelo luphahlazeke.
Emashumini eminyaka amuva nje, ukubala kuye kwashintsha kusuka kumaseva angawodwa kuya kumanethiwekhi amakhulu, ahlanganisiwe emishini asatshalaliswa emhlabeni wonke. Lesi sihloko sizobheka ngokuningiliziwe ukuthi uhlelo olusatshalaliswe luyini, ukuthi luhluke kanjani ohlelweni oluphakathi, izinzuzo kanye nokungalungi kwalo, ukuthi lushintshe kanjani, izinhlobo ezahlukene zezakhiwo ezikhona, lapho lusetshenziswa khona ezinhlelweni zokusebenza zangempela, kanye nezinselele ezizivezayo maqondana nokuxhumana, ukuphepha, ukuphathwa, kanye nokugcinwa kwedatha.
Iyini isistimu esabalalisiwe?
Uhlelo olusatshalaliswe ngokuyisisekelo luyiqoqo lamakhompyutha noma ama-node abambisana ukuhlinzeka ngensizakalo eyodwa ngendlela ehlelekile, njengokungathi angumshini owodwa onengqondo. I-node ngayinye ineprosesa yayo, inkumbulo, kanye nesitoreji, kodwa konke kuxhumana ngenethiwekhi (ngokuvamile i-inthanethi noma inethiwekhi yenkampani) ukuze kwabelwane ngezinsiza futhi kwahlukaniswe umthwalo womsebenzi.
Esikhundleni sokuthembela kuseva eyodwa enkulu ephakathi, umthwalo usatshalaliswa phakathi kwemishini eminingi emincane . Lo mbono uvame ukuqhathaniswa ne-orchestra: insimbi ngayinye (i-node) inengxenye yayo, kodwa lokho okubonwa yizilaleli ukusebenza okukodwa, okuvumelanayo (uhlelo olusatshalalisiwe).
Le ndlela ifanelana kahle nezwe lanamuhla le-big data: ukugcina nokucubungula ulwazi oluningi kakhulu kusebenza kuphela ngokusabalalisa umthwalo womsebenzi emishinini eminingi. Ngakho-ke, ezindaweni zedatha nokuhlaziya kanye ne-Big Data , cishe konke kuncike ezinhlelweni ezisatshalaliswayo: amapulatifomu afana ne-Hadoop, i-Spark, i-Databricks, i-Cloudera, kanye nezinjini zemibuzo ezifana ne-Presto asekelwe kule filosofi.
Isici esibalulekile salezi zinhlelo ukuthi zifihla ubunzima bangaphakathi kumsebenzisi wokugcina . Umuntu osebenzisa iwebhusayithi ye-e-commerce, ibhange eliku-inthanethi, noma isevisi yamafu akaboni amakhulu noma izinkulungwane zama-node, kodwa kunalokho ubona uhlelo lokusebenza "olusebenza nje," yize kukhona ingqalasizinda esakazeke kakhulu ngaphansi.
Umehluko phakathi kohlelo oluhlanganisiwe nohlelo olusatshalaliswe
Kuhlelo oluhlanganisiwe, yonke i-logic, idatha, kanye nokucubungula kugxilwe emshinini owodwa noma kuseva eyinhloko . Uma leyo seva iwa, isevisi ayitholakali kuze kube yilapho ibuyiselwa. Ukwandisa isivinini ngokuvamile kuhilela ukuthenga imishini ebiza kakhulu futhi enamandla, futhi kukhona "iphuzu elilodwa lokwehluleka" elicacile.
Ngokuphambene nalokho, ohlelweni olusatshalaliswe, imisebenzi yabelwana phakathi kwama-node amaningana axhumene . Akukho mshini owodwa obalulekile: uma eyodwa yehluleka, okusele kungaqhubeka nokusebenza futhi kukhokhele lokho kulahlekelwa. Lokhu kwandisa ukubekezelela amaphutha futhi kuvumela ukukhula ngokungeza ama-node amaningi esikhundleni sokukhulisa eyodwa.
Lo mehluko uphinde uthinte indlela esilinganisa ngayo umthamo. Ukusabalala okuvundlile , okuvamile kwezinhlelo ezisatshalaliswe, kuhilela ukwengeza ama-node engeziwe ku-cluster, ukuwabeka "ngokufanayo" ukuze kusatshalaliswe umthwalo kanye nesitoreji.
Ngokombono wezindleko, ngokuvamile kuyabiza kakhulu ukuba namaseva amaningi ajwayelekile asebenza ndawonye kune-superserver eyodwa noma ezimbili ezibiza kakhulu. Ngaphezu kwalokho, ukwehluleka kwe-node encane kuvame ukuba nomthelela omncane kusevisi iyonke, kanti ukwehluleka kweseva enkulu, ephakathi nendawo kungawisa yonke into.
Ingabe izinhlelo ezisatshalaliswe ziyafana nezinsizakalo ezincane?
Nakuba zihlobene kakhulu, azifani ncamashi . Uhlelo olusatshalaliswe luwumqondo obanzi: noma yiliphi isethi yama-node abambisana ngenethiwekhi ukuze anikeze isevisi eyabiwe iwela ngaphansi kwale ncazelo, kungakhathaliseki ukuthi isofthiwe ihlelwe kanjani ngaphakathi kwayo.
Ngakolunye uhlangothi, ukwakheka kwe-Microservices kuyindlela ethile yokuklama izinhlelo zokusebenza ezisatshalaliswayo . Esikhundleni sokudala "i-monolith" eyodwa, uhlelo lokusebenza luhlukaniswe ngezinsizakalo ezincane, ezizimele, ngayinye inomqondo wayo futhi ngokuvamile isizindalwazi sayo. Lawa ma-microservices axhumana omunye nomunye esebenzisa ama-API noma imiyalezo.
Ngakho-ke, ipulatifomu esekelwe kuma-microservices ihlala iyisistimu esatshalaliswe ngoba izingxenye zayo zisakazeke kunethiwekhi futhi zixhunywe yinethiwekhi . Kodwa-ke, kukhona nezinhlelo ezisatshalaliswe ezingalandeli iphethini yama-microservices, njengeqoqo lekhompyutha elihambisanayo, isizindalwazi esisatshalaliswe sakudala, noma inethiwekhi yokwabelana ngamafayela phakathi kontanga.
Izinhlelo ezisatshalaliswe zithuthuke kanjani?
Ezinsukwini zokuqala zokubala kwebhizinisi, kwakuvamile ukuba nezinhlelo ezinkulu, ezihlanganisiwe noma ama-mainframe azenza cishe konke: ukucubungula, ukugcina, ukubika, njll. Ngokuhamba kwesikhathi, kwavela ukwakheka kwezakhiwo zeklayenti-server kanye nezindawo zokugcina idatha ezihlanganisiwe zokuhlaziya ibhizinisi.
Inkinga yayiwukuthi, njengoba idatha ikhula, lezi zindawo zokugcina impahla ezihlanganisiwe zahluleka kokubili ngomthamo nangesivinini . Ukugcina idatha eningiliziwe, yomlando evela emithonjeni eminingi kwakuba kubi kakhulu futhi kuhamba kancane. Izidingo ezintsha zokuhlaziya zazidinga izikhathi zokuphendula ezisheshayo, ubuncane obukhulu, kanye nokucutshungulwa okuhambisanayo.
Yilapho izinhlelo zesimanje ezisatshalaliswayo zingena khona, ikakhulukazi ngokukhuphuka kwe-Big Data kusukela ngawo-2000 kuqhubeke . Nakuba umqondo wokubala okusatshalaliswayo usukela ngawo-1960, amaphrojekthi afana ne-Hadoop kuqala bese kuba yi-Spark (eyadalwa ngo-2009 ukuze kuthuthukiswe ukusebenza nokuguquguquka) enza lo mbono waba yindinganiso ekuhlaziyeni idatha.
Ushintsho lwalusukela ekuzameni ukwenza konke ngethuluzi elilodwa elijwayelekile kuya ekusebenzeni ngezitaki zobuchwepheshe : inhlanganisela yezingxenye ezikhethekile (isitoreji esisatshalaliswayo, izinjini zokucubungula ama-batch kanye nokusakaza, abahleli, amakhathalogi edatha, njll.) ahlanganiswe ndawonye ukuze amboze wonke umjikelezo wokuphila kwedatha.
Isebenza kanjani isistimu esabalalisiwe?
Noma yiluphi uhlelo olusatshalaliswe lungabhekwa njengeqoqo lezingxenye ezilawula isitoreji, ukucubungula, kanye nokuxhumana . I-node ngayinye ithola ingxenye yedatha noma umsebenzi, yenze umsebenzi wayo, bese ihlanganisa imiphumela yayo nesistimu yonke ukuze inikeze impendulo ehlangene.
Ezimweni eziningi, idatha ihlukaniswe ngamabhulokhi, futhi lawa mabhulokhi asatshalaliswa kuma-node ahlukene. Ifayela ngalinye noma irekhodi lingahlukaniswa futhi liphindwe ukuze kube namakhophi angafuneki kumaseva ahlukene. Uma i-node yehluleka, uhlelo lungakha kabusha ulwazi oluvela kuma -replica akhona.
Leli cebo lokuhlukanisa nokuphindaphinda linciphisa kakhulu izikhathi zokufunda nokucubungula , njengoba livumela ukucutshungulwa kwezingcezu ezahlukene ngasikhathi sinye. Ngesikhathi esifanayo, linikeza ukubekezelela amaphutha aphezulu: ukulahlekelwa yi-node eyodwa kuholela ekunciphiseni okuncane kwamandla, hhayi enhlekeleleni yomhlaba wonke.
Kodwa-ke, wonke lomlingo uza ngezindleko zobunkimbinkimbi: ukuphatha, ukumisa, kanye nokuqapha amaqoqo asakazwe akuyona into encane . Kudinga ukuhlanganisa izibuyekezo, ukuqapha isimo sama-node, ukuphatha ukusatshalaliswa kwedatha lapho usayizi weqoqo ushintsha, kanye nokuxazulula izinkinga zokungaguquguquki phakathi kwamakhophi.
Ukwakhiwa kwezinhlelo ezisatshalaliswayo
Kunezindlela eziningana zokwakha zokuhlela uhlelo olusatshalalisiwe, ngalinye linezinzuzo zalo kanye namacala okusetshenziswa. Ezivame kakhulu zihlanganisa ama-topology ahlukene okuxhumana kanye nokusatshalaliswa kwemithwalo yemfanelo phakathi kwama-node.
Enye yezakhiwo ezidumile kakhulu imodeli yeklayenti-iseva . Kulo modeli, iseva eyodwa noma ngaphezulu inikeza izinsiza (idatha, izinsizakalo, amafayela), futhi amaklayenti enza izicelo futhi asebenzise lezo zinsiza. Kufana nomtapo wolwazi: umtapo wolwazi (iseva) uphatha izincwadi, futhi abasebenzisi (amaklayenti) bayazicela.
Ngakolunye uhlangothi kukhona ukwakheka kontanga , lapho kungekho node ephakathi elawula konke. Umhlanganyeli ngamunye usebenza njengeklayenti kanye neseva, abelana ngezinsiza nabanye. Lena imodeli evamile yamanethiwekhi amaningi okwabelana ngamafayela kanye namanye ama-cryptocurrencies.
Okunye okuphawulekayo yizakhiwo ezigxile kusevisi kanye nezama-microservices , lapho uhlelo lokusebenza lwakhiwe khona ngezinsizakalo eziningi ezisatshalaliswayo eziveza izixhumi ezichazwe kahle. Isevisi ngayinye ingafakwa, ilinganiswe, futhi ibuyekezwe ngokuzimela, okuhlinzeka ngokuguquguquka okukhulu kokuvela kwesistimu.
Kuzo zonke izimo, isihluthulelo sisekutheni ama-node ahlanganiswa futhi avumelaniswa kanjani: ukuvumelana kwemali, ukubambezeleka, ukwehluleka okungaphelele, kanye nokuvumelana kwedatha kumele kuphathwe , kuyilapho kugcinwa ulwazi lomsebenzisi olubushelelezi noluhambisanayo.
Izinzuzo zamasistimu asabalalisiwe
Phakathi kwezizathu ezenza ukuthi izinhlelo ezisatshalaliswe zibe yindinganiso emikhakheni eminingi kangaka, izinzuzo eziningana ezicacile kakhulu ziyavela ezihlobene nokusebenza, ukutholakala kanye nokukhula.
Enye yezinzuzo ezibonakala kakhulu ukusebenza okuthuthukisiwe . Ngokuvumela imishini eminingi ukuthi isebenze ngesikhathi esisodwa ezingxenyeni ezahlukene zomsebenzi, izikhathi zokuphendula ziyancishiswa futhi kusekelwa umthwalo womsebenzi omningi kakhulu. Lokhu kubalulekile ezinhlelweni ezibalulekile njengebhange eliku-inthanethi, i-e-commerce, kanye nezinsizakalo zesikhathi sangempela.
Enye inzuzo enkulu ukutholakala okuphezulu . Ngokusabalalisa kokubili umthwalo womsebenzi kanye nedatha kuma-node amaningi, uma eyodwa yehluleka, uhlelo lungaqhubeka nokusebenza ngokuthembela kwamanye. Lokhu kuqina kubalulekile lapho isikhathi sokungasebenzi siholela ngqo ekulahlekelweni kwezezimali noma ulwazi olubi lomsebenzisi.
Ukusabalala nakho kuyisihluthulelo samandla: izinhlelo ezisatshalaliswe zingakhula ngokungeza ama-node kunethiwekhi ngaphandle kokuphazamisa isevisi. Lokhu kuzivumela ukuthi zivumelane nesidingo esiphezulu, ukukhula kwebhizinisi okuqhubekayo, noma izinguquko kuvolumu yedatha, kugwenywe isidingo sokuvala imisebenzi ukuze kuthuthukiswe kuseva enamandla kakhulu.
Ngaphezu kwalokho, zinikeza ukuguquguquka okukhulu ekuphathweni kwezinsizakusebenza . Imisebenzi ethile ingabekwa phambili, amandla engeziwe angabelwa izinqubo ezibalulekile, noma izinsizakalo ezintsha zingafakwa kuma-node athile. Leli khono lokulungisa kahle libaluleke kakhulu ezindaweni ezishintshashintshayo kakhulu.
Izinkinga kanye nezingozi zezinhlelo ezisatshalaliswe
Akuzona zonke izinzuzo: ukusatshalaliswa kwethula izinkinga ezintsha ezingaveli (noma ezingaveli njalo) ezinhlelweni ezihlanganisiwe. Ukuklama nokusebenzisa lezi zakhiwo kuhilela ukubhekana nezinselele ezithile.
Okokuqala, kukhona ubunzima bokuxhumana . Uma usebenza kumanethiwekhi angempela, kufanele ubhekane nokubambezeleka okuguquguqukayo, umkhawulokudonsa olinganiselwe, ukulahleka kwephakethe, kanye nokungafani phakathi kwama-node. Ukuxhumanisa izinqubo ezabelana ngedatha kunethiwekhi yonke ngaphandle kokuvimba uhlelo noma ukudala ukungahambisani akuyona into encane.
Enye inkinga ebalulekile ukwehluleka namaphutha . Endaweni esatshalaliswe, cishe akunakugwenywa ukuthi i-node, idiski, noma isixhumanisi senethiwekhi sizohluleka ngesinye isikhathi. Ngakho-ke, izindlela eziqinile zokuthola ukwehluleka, ukululama okuzenzakalelayo, ukuzama kabusha ukusebenza, kanye nokusabalalisa kabusha imisebenzi nedatha okunamandla kubalulekile.
Ukuphepha nakho kuba nzima kakhulu: lapho ama-node eba maningi, indawo yokuhlasela iba nkulu. Izinhlelo ezisatshalaliswayo zisengozini enkulu yokuhlaselwa njengokwenqatshwa kwenkonzo, ukufakwa kwekhodi, ukuvinjelwa kokuxhumana , noma ukufinyelela okungagunyaziwe kuma-node angavikelwe kahle.
Okokugcina, ukuphatha nokuphatha kudinga kakhulu. Ukulungisa, ukuqapha, nokugcina iqembu elisakazeke ngokwendawo elakhiwe ubuchwepheshe obungafani kudinga amathuluzi amahle, izinqubo ezivuthiwe, kanye namaqembu obuchwepheshe anolwazi oluthile kulezi zinhlobo zezindawo.
Izinhlelo zokusebenza zangempela zezinhlelo ezisatshalaliswayo
Ukuba khona kwezinhlelo ezisatshalaliswe empilweni yansuku zonke kusakazeke kakhulu kangangokuthi kunzima ukucabanga ngezinsizakalo zedijithali zesimanje ngaphandle kwazo. Imikhakha eminingi ebalulekile ithembele kulolu hlaka ukuze isebenze ngokwethembeka.
Ezweni lewebhu, isibonelo, izinhlelo zokusebenza ezinkulu ze-e-commerce kanye nezemidiya yezenhlalo zisebenzisa izinhlelo ezisatshalaliswe ukuze zikhonze izigidi zabasebenzisi ngesikhathi esisodwa. Amapulatifomu afana ne-Amazon ne-Alibaba asakaza izicelo kuzo zonke izikhungo zedatha emhlabeni wonke futhi asekela ukwandiswa kwazo ngama-database asatshalaliswe kanye namanethiwekhi okulethwa kokuqukethwe (ama-CDN).
Amanethiwekhi okuxhumana ngocingo kanye ne-inthanethi athembele ezingqalasizinda ezisatshalaliswe ezithumela izingcingo, imiyalezo, kanye namaphakethe edatha ngama-node amaningi aphakathi. Lokhu kuvumela ukuxhumana ukuthi kugcinwe amazinga afanele okubambezeleka nokuthembeka ngisho noma ingxenye yenethiwekhi ihlangabezana nokuphazamiseka.
Umkhakha wezezimali kanye nowamabhange ungesinye isibonelo esihle: izinhlelo zokukhokha, ama-ATM, ukuhweba, kanye nokubhanga ku-inthanethi kuncike kudathabheyisi nezinsizakalo ezisatshalaliswayo eziphinda ulwazi kuzo zonke izifunda, zisebenzisa izindlela eziqinile zokubethela kanye nokuqinisekisa, futhi zisekela ukuthengiselana okusakazeke ngokwendawo ngenkathi kunciphisa ingozi yokwehluleka.
Emkhakheni we-Big Data kanye nokuhlaziywa okuthuthukisiwe, izinhlelo zokucubungula ezisatshalaliswayo zivumela ukusebenza ngamanani amakhulu amarekhodi: amalogi eseva, idatha yenzwa, izinkundla zokuxhumana, ukuthengiselana, njll. Ubuchwepheshe obufana ne-Hadoop Distributed File System (HDFS) noma i-Spark busabalalisa isitoreji kanye nokubala kuma-node amaningi ukuqinisekisa izikhathi zokucubungula ezifanele.
Izinhlelo zedathabheyisi ezisatshalalisiwe
Ama-database asatshalaliswa ayisibonelo esikhethekile nesibaluleke kakhulu ngaphakathi kwezinhlelo ezisatshalaliswayo. Esikhundleni sokugcina yonke idatha kuseva eyodwa, isatshalaliswa kuma-node amaningana , avame ukutholakala ezifundeni ezahlukene zezwe, okugcina umbono onengqondo ohlangene wombuzo.
Leli cebo livumela ukukhuliswa kokubili kumthamo wokugcina kanye nokusebenza kokufunda/ukubhala. Ama-node noma izifunda ezintsha zingangezwa njengoba isidingo sikhula , futhi izindlela zokuhlukanisa kanye nokuphindaphinda zisingatha ukusatshalaliswa kabusha kolwazi ngokuzenzakalelayo.
Enye yezinselele ezinkulu ukugcina ukuvumelanisa idatha kanye nokuvumelana phakathi kwamakhophi. Lokhu kufezwa kusetshenziswa ama-algorithms okuvumelana afana ne-Paxos noma i-Raft, okuqinisekisa ukuthi imisebenzi isetshenziswa ngokulandelana okuhambisanayo kuwo wonke ama-node ngaphakathi kweqembu lokukopisha.
Kuye ngohlobo lohlelo lokusebenza, amanye ama-database abeka phambili ukutholakala nokubekezelelana kwezingxenye zenethiwekhi kunokuhambisana okuqinile, asebenzisa amamodeli afana nokuvumelana ekugcineni . Kwezinye izimo, ukuphindaphinda okuhambisanayo kusetshenziselwa ukugcina ukuvumelana okuqinile, okulahla ukubambezeleka okuthile ngokushintshana nobuqotho bedatha obukhulu.
Amapulatifomu amakhulu e-e-commerce kanye nezinsizakalo zamafu zihlanganisa izizindalwazi ezisatshalaliswe nezinhlelo zokugcina isikhashana ukuze kunikezwe okuqukethwe okune-latency ephansi futhi kusingathwe ukwanda kwethrafikhi. Isibonelo esivamile sesitoreji esisatshalaliswe esigxile ekuthembekeni nasekubekezeleleni amaphutha yi-Amazon S3, ephinda idatha kumaseva amaningi ngaphakathi kwesifunda.
Ukubala okuhambisanayo kanye nokusebenza okuphezulu ezinhlelweni ezisatshalalisiwe
Enye indawo lapho izinhlelo ezisatshalaliswa zikhanya khona yi -high-performance parallel computing (HPC) . Esikhundleni sokucubungula inani elikhulu ledatha ngokulandelana emshinini owodwa, izibalo zisatshalaliswa kuwo wonke amaqoqo amakhulu noma izinkulungwane zama-node.
Kulezi zigaba, i-node ngayinye yenza ingxenye yenkinga, futhi ngokusebenzisa amasu okuhlanganisa ahlelwe kahle, imiphumela engaphelele iyahlanganiswa ukuze kutholakale umphumela wokugcina . Lokhu kwenza kube nokwenzeka ukubhekana nokulingisa okuyinkimbinkimbi kwesayensi, ukumodela isimo sezulu, ukuhlaziywa kwezezimali okuthuthukisiwe, noma ukucutshungulwa kwezithombe ezinkulu zezokwelapha ngesivinini esingacabangeki ngomshini owodwa.
Ukuze kufezwe lokhu kusebenza kahle, kusetshenziswa ama-algorithm ahambisanayo aklanyelwe ngqo ukusabalalisa umthwalo nokunciphisa ukuxhumana phakathi kwama-node . Amasu anjengokunamathelana kwe-CPU noma ukwenza ngcono izakhiwo ze-NUMA asiza ukuthuthukisa ukusebenza ngokulungisa indlela izinqubo nedatha ezibekwa ngayo kwimemori kanye nama-processor.
Kubuhlakani bokwenziwa kanye nokufunda okujulile, i-distributed computing ivumela ukuqeqeshwa kwamanethiwekhi amakhulu e-neural ngokusabalalisa idatha namamodeli kuma-GPU amaningi namaseva . Uhlelo luhlanganisa ama-gradients kanye nezibuyekezo zamapharamitha ukuze ukuqeqeshwa kuqhubeke ngesikhathi esifanayo ngaphandle kokwephula ukuhambisana kwemodeli.
Ifu lithuthukise le ndlela ngokunikeza i-HPC njengesevisi (i-HPCaaS) , ukuze amabhizinisi amancane namaqembu akwazi ukuqasha okwesikhashana amaqoqo amakhulu ukuqeqesha amamodeli noma ukuqhuba ukulingisa okujulile, ngaphandle kokuthenga nokugcina yonke leyo ngqalasizinda ngokuphelele.
Izinhlelo ezisatshalaliswa kubuchwepheshe bansuku zonke
Ngaphandle kwezikhungo zedatha, izinhlelo ezisatshalaliswe ziyingxenye yokuphila kwansuku zonke kwanoma ubani oxhumana nobuchwepheshe. Ukuba khona kwazo kuvamile kangangokuthi asiziboni kahle.
Izinsizakalo ze-imeyili, amapulatifomu okuthumela imiyalezo esheshayo, kanye namanethiwekhi omphakathi asebenza ezingqalasizinda ezisatshalaliswayo eziphinda idatha yomsebenzisi emhlabeni wonke . Ngenxa yalokhu, singafinyelela imiyalezo yethu kunoma iyiphi idivayisi, ngokubambezeleka okuphansi futhi, ngokuvamile, ngaphandle kokuphazamiseka okubonakalayo.
Amanethiwekhi okwabelana ngamafayela kontanga ngontanga ayisibonelo esinye: esikhundleni sokulanda kusuka kuseva eyodwa, ifayela liyahlukaniswa futhi likhonzwa kusuka kontanga abaningi , lapho umuntu ngamunye ohlanganyelayo esebenza khona ngesikhathi esisodwa njengeklayenti kanye neseva, okuthuthukisa ukuqina kwenethiwekhi kanye nokusebenza kwayo.
Emkhakheni we- Internet of Things (IoT) kanye nama-smart grid, izigidi zezinzwa namadivayisi athumela idatha kumapulatifomu asatshalaliswayo acubungula ulwazi ngesikhathi sangempela ukuze athuthukise ukusetshenziswa kwamandla, enze ngokuzenzakalelayo izakhiwo, noma axhumanise izimoto ezixhunyiwe.
Futhi, vele, amapulatifomu amakhulu e-cloud computing njenge-AWS, i-Microsoft Azure, noma i-Google Cloud ayisibonelo esisobala kakhulu sesistimu esatshalaliswe: ahlanganisa izikhungo zedatha ezifundeni ezahlukene, anikeza izinsiza ezidingekayo, futhi avumela izinkampani ukuthi zisebenzise izinhlelo zazo emhlabeni wonke ngokuchofoza okumbalwa kanye nekhadi lesikweletu.
Ngingazi kanjani ukuthi hlobo luni lwesistimu esakazwayo engiludingayo?
Lapho ukhetha ikhambi elithile, ayikho iresiphi eyodwa: ukwakheka kohlelo olusatshalalisiwe kumele kuhambisane nesimo senhlangano , izinhloso zayo, kanye nokuvuthwa kwayo kwezobuchwepheshe.
Kungcono ukuqala ngokuhlaziya ivolumu yedatha yamanje nelindelekile . Ukucubungula amarekhodi ayizigidi ezimbalwa ngosuku akufani nokuphatha ukusakazwa kwedatha okuqhubekayo, kwesikhathi sangempela okuvela kumadivayisi e-IoT asatshalaliswa emhlabeni wonke.
Kubalulekile futhi ukucabangela isabelomali esitholakalayo kanye nesu lokukhulisa . Ezinye izinkampani zingakhokhela amaqembu azinikele kanye nabasebenzi abakhethekile, kanti ezinye zizothembela cishe ngokuphelele kumasevisi efu aphethwe ukuze kuncishiswe ubunzima bokusebenza.
Kubalulekile futhi ukucabangela isidingo esiphezulu, izikhathi zomsebenzi ophansi, kanye nemikhawulo yesikhathi sokucubungula . Uhlelo oludinga ukuphendula ngama-millisecond luzoba nezidingo ezihlukile kunelo oluklanyelwe ukucutshungulwa kweqembu ebusuku.
Ukuchaza lezi zici kusukela ekuqaleni kusiza ekwakheni isakhiwo esivumelanayo okulula ukusiphatha futhi esingathwali ukumangala. Namuhla, ngisho nezinhlangano ezincane zingafinyelela umthamo wokubala osabalalisiwe owawutholakala ezinkampanini ezinkulu kuphela , uma nje zinolwazi lobuchwepheshe oludingekayo kanye nobuhlakani bebhizinisi ukuze zilusebenzise.
Izinhlelo ezisatshalaliswe zishintshe kusukela esixazululweni esikhethekile zaba umgogodla wezinsizakalo eziningi zedijithali. Amandla azo okusabalalisa umthwalo, ukubekezelela amaphutha, ukukala ngokuvundlile, nokuphatha amanani amakhulu edatha kuzenza zibe yingxenye ebalulekile kunoma iyiphi inhlangano efuna ukuncintisana endaweni exhumene kakhulu, edinga kakhulu, futhi encike kubuchwepheshe.
