Ibhulogi Yendwangu Ye-Microsoft: Umhlahlandlela Othuthukisiwe Wezici Ezintsha Nokusetshenziswa

Informatec Digital » Izinsiza » Ibhulogi Yendwangu Ye-Microsoft: Umhlahlandlela Othuthukisiwe Wezici Ezintsha Nokusetshenziswa

Uma ubulokhu usebenza ngedatha ku-ecosystem ye-Microsoft isikhathi eside, uzobe ukuphawulile lokho I-Microsoft Fabric isibe yisikhungo esiyinhloko sokuhlaziya kwesimanjeUkuhlanganisa konke okwakusakazeke ngaphambili ku-Power BI, i-Azure Synapse, i-Data Factory, kanye nezinye izinsizakalo zibe yipulatifomu eyodwa. Kulesi sihloko, sizobheka ngokuningiliziwe izici zakamuva, imephu yomgwaqo, kanye nemiphumela esebenzayo ye-Fabric ekuhlaziyeni, ekubuseni, e-AI, kanye nokubala kwesikhathi sangempela.

Umqondo uwukuthi, uma usuqedile ukufunda, uzoba nokuqonda okucacile Yini enikezwa yiMicrosoft Fabric namuhla, iyaphi, futhi iyithinta kanjani i-data architecture? yenhlangano yakho: amakhono okuphatha nge-Purview, ukuzenzekela ngama-API kanye ne-Git, izici ezintsha ze-AI, ubuhlakani besikhathi sangempela, ukuphepha ku-OneLake, ukufuduka, ukusebenza nokunye okuningi.

I-Microsoft Fabric kanye noHlaka Lokwamukelwa Kwamafu: ukubusa, izindleko, kanye nokwenza izinto ngokuzenzakalela

Ngaphakathi kwe-Cloud Adoption Framework (CAF), iMicrosoft ibilokhu ishicilela uchungechunge lweziqondiso zokuvumelanisa Umklamo we-Microsoft Fabric onezinsika zokuklama zezindawo zokufikelaIngxenye yokugcina yalolu chungechunge ikhuluma ngezindawo ezintathu ezibalulekile: ukubusa, ukwenza ngcono amandla, kanye nokwenza izinto ngokuzenzakalela/ama-DevOps.

Esigabeni sokuphatha, Izindleko zomthamo wendwangu zivezwa nge-Azure Cost ManagementLokhu kukuvumela ukuthi uhlanganise idatha yokusetshenziswa kwe-F capacity (nezinye i-SKU) kumadeshibhodi ezindleko ze-Azure, izexwayiso, kanye nokuhlaziywa kwemali. Akukhona nje ukubona ibhili: ungabheka lolu lwazi ngamathegi, okubhaliselwe, noma amaqembu ezinsiza ukuze uqonde ukuthi ubani osebenzisa ini nokuthi kungani.

Ukwakhiwa kobuchwepheshe be-Fabric kanye nemodeli yebhizinisi kuvumela amakhasimende lawula kahle ukusetshenziswa kwamandla ukuze kuthuthukiswe izindleko zokusebenzaEmpeleni, lokhu kuhunyushwa kube yizibambo eziningana:

• Ukwenyuka kwamakhono F: ukwandisa noma ukunciphisa amandla ngokusekelwe eziqongweni ezibikezelwayo, njengemikhankaso yokuthengisa, izinsuku zokugcina zokubalwa kwezimali, noma imithwalo yedatha emikhulu.
• Amakhono okumisa isikhashana nokuqalisa kabushaAbaphathi bezindwangu bangamisa umthamo we-F uma ungadingeki (isibonelo, ebusuku noma ngezimpelasonto ezindaweni ezingezona ezibucayi) ukuze banciphise izindleko zokubala.
• Ukuvikelwa kokuqhumaAmapharamitha amabili ezingeni lomthamo alawula indlela imisebenzi yangemuva ephathwa ngayo ukuze ivimbele ukugcwala kwemvelo: I-Background Rejection Threshold kanye ne-Background Recovery Threshold.
• Ukubhukha umthamoKungenzeka ukubhukha umthamo wesikhathi esibekiwe ukuze uthole izaphulelo, uma nje ukusetshenziswa okulindelekile kuhlelwe kahle.

Emkhakheni wokuphathwa kwedatha, i-Fabric ithembele ku I-Microsoft Purview njengengxenye ebalulekile yokwenza uhlu, uhlu lozalo, kanye nokuhlelaKusukela ku-Purview kungenzeka ukusebenzisa amalebula okuzwela, ukubona uhlu phakathi kwemvelaphi, ukuguqulwa kanye nokusetshenziswa, ukuqinisekisa izimpahla zedatha noma ukuthuthukisa imikhankaso yekhwalithi yedatha, konke lokhu ngenkathi kuhlanganiswa izakhi ze-Fabric (Lakehouse, Warehouse, KQL, Power BI, njll.).

Isigaba esithi "I-Platform Automation kanye ne-DevOps" sakhiwe ngokuyinhloko ngokuphathelene Ukuhlanganiswa ne-Git, amapayipi okuhambisa, nama-REST APIIzindawo zokusebenza zendwangu zingaxhunyaniswa nezindawo zokugcina ze-Git ukuze abathuthukisi abaningi bakwazi ukubambisana kuphrojekthi efanayo yobunjiniyela bedatha, isayensi yedatha, noma yokuhlaziya ngesikhathi sangempela ngaphandle komsebenzi ohambisanayo.

Amapayipi e-Git kanye nokusetshenziswa avumela ukuhlela intuthuko, ukuhlolwa, kanye nemijikelezo yokuphila yokukhiqiza ngendlela elawulwayoUkukhushulwa kokuqukethwe, ukulandelela ushintsho, ukuvumelanisa izinto ezidaliwe, kanye nokuphathwa kwenguqulo. Kusukela ngoMashi 2025, lawa mapayipi asekele (ngokujwayelekile noma esimweni sokubuka kuqala) izinhlobo eziningi zezinto ezifana nama-trigger, amadeshibhodi, ukugeleza kwedatha, amapayipi edatha, ama-datamart, izindlu zamachibi, izindawo zokugcina impahla, i-KQL, ama-notebook, imibiko ebhalwe ngamakhasi, izinhlelo zokusebenza zenhlangano, njll.

Okwamanje, i-Fabric iyadalula isethi ebanzi kakhulu yama-REST API ongakwenza ngazo ngokuzenzakalelayo cishe noma yimuphi umsebenzi wokuphatha noma wokusabalalisa: ukudalwa kwezindawo zokusebenza namafolda, ukufuduka kwezinto ngobuningi, ukuphathwa kokuxhumeka namasango, ukukhipha izincazelo, ukusebenza kwamapayipi noma ama-notebook, phakathi kwezinye izinto.

Izibuyekezo ezibalulekile ku-analytics, i-AI, kanye nokuphathwa kwe-Microsoft Fabric

Ukuvela kwe-Fabric kuphawulwa yinani elikhulu lezici zokubuka kuqala kanye namakhono amasha Zithinta cishe zonke izindawo zesikhulumiBese zihlanganiswa futhi zixoxwe ngendlela ehlelekile ukuze kubonakale yonke imininingwane.

Imisebenzi ye-AI ku-Data Warehouse kanye nemisebenzi ye-multimodal

Endaweni yedatha yobudlelwano, i-Fabric Data Warehouse ihlanganisa I-AI isebenza ngqo ku-T-SQL (ukuhlola kuqala). Lokhu kuvumela:

• Hlela umbhalo ngokwezigaba noma uhlukanise ngezigaba.
• Hlaziya imizwa.
• Khipha ulwazi oluhlelekile embhalweni wamahhala.
• Humusha umbhalo phakathi kwezilimi.
• Uhlelo lolimi olufanele.

Inhloso ukuthi Asikho isidingo sokushiya umongo we-SQL ukuze ucebise idatha nge-AILe misebenzi ihlanganiswe nokusekelwa okujwayelekile kwe-multimodal emisebenzini ye-AI ye-Fabric, manje engakwazi ukucubungula izithombe (JPG/JPEG, PNG, GIF, WebP), ama-PDF namafomethi ombhalo avamile, kanye nokufakwayo ngendlela yezindlela zamafayela.

Izinsiza ezifana aifunc.load yokufaka amafolda ematafuleni Ngezinketho ze-prompt kanye ne-schema, i-`aifunc.list_file_paths` ikuvumela ukuthi uphindaphinde izindlela zefayela, futhi i-`ai.infer_schema` iphetha ngokuthi ama-schema ahambisana ne-`ai.extract`. Konke lokhu kukuvumela ukuthi uguqule futhi ucebise idatha ngaphandle kokwakha amapayipi ayinkimbinkimbi kusukela ekuqaleni.

I-OneLake, i-Delta, i-Apache Iceberg kanye nokufinyelela kumapulatifomu amaningi

Engqimbeni yokugcina izinto, i-OneLake isalokhu iyichibi elihlangene lapho kuhlala khona yonke i-Fabric. Enye yezinguquko ezinkulu yikhono lokwenza Ukuveza amatafula eDelta Lake njengamatafula e-Apache Iceberg ngaphandle kokuhambisa noma ukukopisha idatha, ukuze izinjini ezihambisana ne-Iceberg zikwazi ukufunda ngqo okuku-OneLake.

Ngaphezu kwalokho, i-OneLake manje iyasekela Idatha ye-Apache Iceberg ebhalwe ngqo yi-Snowflake, isetshenziswe ku-Fabric ngokufinyelela okungenakhophiLeli qhinga liqinisa ukusebenzisana: Indwangu ayifuni ukuba yi-silo, kodwa ifuna ukuba "isikhungo" sedatha ezinye izinjini ezingasisebenzisa ngaphandle kokuphinda isitoreji.

Ngendlela efanayo, i-OneLake iyavuma izinqamuleli ze-Azure Blob StorageI-OneDrive kanye ne-SharePoint, futhi yandisa ukuphepha ngezindima zokufinyelela, ukuphepha kwefolda, umugqa kanye nezinga lekholomu, kanye nemodeli yokuphepha abantu besithathu abangayihlonipha ngenxa yokwandiswa kwemodeli yenjini egunyaziwe.

Iphuzu elilodwa elibaluleke kakhulu yi-federation yekhathalogi ye-OneLake ku-Azure Databricks, evumela Ukufinyelela okungakopishiwe kusuka ku-Unity Catalog kuya kumathebula e-OneLakeNgale ndlela, i-OneLake ihlala ingumthombo weqiniso, kodwa i-Databricks ingabuza idatha ngqo, ivumelanise i-metadata kuphela.

Idathabheyisi ye-SQL ku-Fabric: ukusebenza, ukuphepha, kanye nokwenza i-virtualization

Isizindalwazi se-SQL se-Fabric sithola amakhono aso: Izinketho ze-ALTER DATABASE SETHIUsekelo lokuqoqa kanye nokufaka umbhalo ogcwele ekubukeni kuqala. Ezingeni le ukusebenza kwesizindalwazi Futhi maqondana nezindleko, kunezintuthuko eziningana:

• Ukuqiniswa kwenkomba okuzenzakalelayo ukunciphisa isitoreji, i-I/O nokuthuthukisa izikhathi zokubuza ngaphandle kokuhlela imisebenzi yokulungisa.
• Umkhawulo ophezulu we-vCores ukulawula ukusetshenziswa kwezinsizakusebenza zokubala (ama-vCores angu-4 noma angu-32), ezenzelwe ukuvimbela umthwalo womsebenzi ukuthi ungadli kakhulu umthamo owabiwe.
• Amachibi e-SQL Angokwezifiso ezinikeza abaphathi bezindawo zokusebenza ukulawula okuningiliziwe kokwabiwa kwezinsiza kanye nokuhanjiswa kwemibuzo ngegama lesicelo.

Iphinde inikwe amandla Ukwenziwa kwedatha ibe yi-virtualization ku-database ye-SQL, okuvumela ukubuza idatha yangaphandle egcinwe ku-OneLake nge-T-SQL, ukuhlanganisa amafayela ngamafomethi ajwayelekile namathebula obudlelwano bendawo kusetshenziswa ama-joins, ngaphandle kokungenisa idatha ngokomzimba.

Ngokuphathelene nokuphepha, i-SQL database iyasekela Isixhumanisi Sangasese ezingeni lomqashi (ukubuka kuqala)Lokhu kwenza kube lula ukuhambisa ithrafikhi yedatha ngendlela eyimfihlo nelawulwayo, kuhlanganiswa nokucushwa kwenethiwekhi okuphephile kwe-Fabric.

Ubuhlakani Besikhathi Sangempela, Indlu Yemicimbi, Ukusakaza Kwemicimbi kanye Nokucutshungulwa Kwemisebenzi

Imojula ye-Real-Time Intelligence (RTI) isibe ngenye yezinto ezibalulekile ezihlukanisa i-Fabric. I-Eventhouse kanye ne-Eventstream ziyahlangana gwinya, sebenza futhi uqalise imicimbi yesikhathi sangempela kusuka kuzo zonke izinhlobo zemithombo, futhi i-Activator ihlela izenzo ezibangelwa ngaphansi kwezimo ezithile.

Phakathi kwezici ezintsha ezinamandla kakhulu yilezi:

• Ukutholwa kwe-Anomaly ngaphandle kwekhodi ngokukhetha imodeli okuzenzakalelayo, isikhombikubona esilula kanye nezaziso eziguquguqukayo.
• Imicimbi Yebhizinisi, ethwebula izikhathi ezibalulekile zebhizinisi ezikhiqizwe ku-User Data Functions kanye nama-Notebook, futhi ikuvumele ukuthi usebenzise izexwayiso, i-logic yangokwezifiso, ukugeleza, amamodeli e-AI noma imisebenzi ye-Spark.
• Ukuhlanganiswa Kwemisebenzi Yedatha Yomqalisi-Umsebenzisiukuze imisebenzi edalwe ku-Fabric ikwazi ukucubungula imicimbi evela kunoma yimuphi umthombo, okuhlanganisa imicimbi yangaphakathi evela epulatifomu uqobo kanye ne-OneLake.
• Amandla okucubungula imicimbi nge-SQL (uMsebenzi we-SQL ku-Eventstream), okuvumela ukuguqulwa kokugeleza kwesikhathi sangempela nge-syntax ye-SQL eyaziwayo.

Izixhumi ezifanele ziyangezwa njenge I-Cribl (kokungenisa amalogi kanye ne-telemetry kusuka emithonjeni eminingi), isixhumi esine-Solace PubSub+, kanye nokusekelwa kokusakaza ngamanethiwekhi azimele nge-Azure Virtual Network, i-VPN, i-ExpressRoute noma ama-endpoint azimele.

Ngezinhlelo zedatha kanye nezinkontileka, i-Eventstream yethula i- Ukubhaliswa kwe-Schema okuchaza futhi kuqinisekise izinhlelo zemicimbi zamapayipi aqinile, kanye nokusekela i-Confluent Schema Registry ukuxhumana ne-Kafka ku-Confluent Cloud ngenkathi kuhlonipha izinkontileka ezikhona.

Amakhono okushayela indiza kanye ne-AI kuyo yonke ipulatifomu

I-Copilot in Fabric isiyatholakala emhlabeni jikelele, futhi ikhona ku I-Power BI, i-Data Factory, i-Data Science kanye nobunjiniyela bedatha kanye nokubhala imibuzo ye-KQLNgaphezu kwalokho, amakhono athile afakiwe:

• Umshayeli we-Dataflow Gen2 (Idatha Yokuthola Yesimanje), okusiza ekungeniseni nasekuguquleni idatha ngemiyalelo yolimi lwemvelo.
• Umshayeli wendiza we-Data Warehouse (ingxoxo), kufinyeleleka ngenkinobho eribhoni ukuze kusheshiswe imisebenzi yokugcina impahla ngebhokisi lengxoxo.
• Ikhophi ye-endpoint ye-SQL analytics, ekhiqiza futhi ithuthukise imibuzo ye-SQL kusukela ezincazelweni zebhizinisi.
• Umshayeli wendiza osizayo kuma-notebook ngolwazi lomongo wendawo yokusebenza, i-lakehouse, isakhiwo se-notebook kanye nendawo yokusebenza, okwazi ukukhiqiza ikhodi yezinyathelo eziningi, ukulungisa kabusha, ukufingqa ama-notebook ayinkimbinkimbi kanye nokuhlonza amaphutha ngenketho ethi "Lungisa nge-Copilot".
• Ukuqedela ngokuzenzakalela okusemgqeni (ukuqedela ikhodi esemgqeni) kanye nokuqedela ikhodi esemgqeni ye-Notebook Copilot (ukubuka kuqala), ukubhala i-Python ngokushesha futhi ngamaphutha ambalwa.

Ngaphezu kwalokho, isisekelo sobuchwepheshe be-AI siyandiswa nge Amathuluzi Okwakhiwa Kwama-Foundry ahlanganisiwe (i-Azure OpenAI, ulimi lwe-Azure, umhumushi we-Azure), ama-plugin e-OpenAI e-Eventhouse (ai_embed_text kanye ne-ai_chat_completion) kanye nochungechunge lwama-ejenti nama-ejenti edatha avumela ezinye izinhlelo zokusebenza, kufaka phakathi i-Copilot Studio, ukuthi zisebenze kudatha ye-Fabric ngendlela ehlelekile.

Ama-Ejenti Edatha Yendwangu, i-MCP kanye namathuluzi onjiniyela

Ukwethulwa kwendwangu ama-ejenti edatha akwazi ukuhlela ukufinyelela kwedatha namathuluzi Kuma-ejenti e-AI, ane-Python SDK kanye nokuhlanganiswa okuqondile ne-Microsoft Copilot Studio. Lokhu kwenza kube lula ukwakha abasizi bengxoxo abasebenza nedatha yebhizinisi elawulwa ku-Fabric.

Ngokuhambisanayo, i- I-Model Context Protocol (MCP) Iba yingxenye ebalulekile yokuhlanganiswa phakathi kwama-ejenti e-AI kanye nezinsizakalo ze-Fabric. Kukhona amaseva e-MCP azinikele e-Activator kanye ne-Eventhouse, kanye ne-Fabric MCP egxile ekuthuthukisweni e:

• Ivumela abasizi be-AI ukuthi bakhiqize ikhodi nokuqukethwe kwezinto zendwangu.
• Ihlangana namathuluzi okuthuthukisa njenge-VS Code kanye ne-GitHub Codespaces.
• Iveza amathuluzi okubonisana nokusebenza ngedatha yesikhathi sangempela ku-Eventhouse.

Ngomsebenzi wansuku zonke womthuthukisi, kunezingcezu eziningana ezibalulekile okufanele zigqanyiswe, okuhlanganisa imvelo yokuthuthukiswa: Isandiso se-MSSQL se-VS Code ngokusekelwa kwesizindalwazi se-Fabric SQL, i-Microsoft ADO.NET Driver kanye ne-ODBC Driver ye-Fabric Data Engineering (uxhumano ne-Spark SQL nge-Livy), kanye nesixhumi se-Spark sedathabheyisi ye-SQL esenza kube lula ukufinyelela okuqinisekisiwe kusuka kudathabheyisi ye-Spark kuya ku-SQL ku-Azure kanye ne-Fabric.

Iphinde ivela I-CLI Yendwangu, itholakala njengomsebenzi ohlanganisiwe ku-Azure DevOps, ekuvumela ukuthi uphathe izindawo zokusebenza, izinto, kanye nokusetshenziswa ngokuzenzakalelayo ngaphandle kokufaka amathuluzi angaphandle ngesandla.

I-Data Factory, ukufuduka kwedatha kanye nokuhlelwa kwendwangu

Isendlalelo sokuhlanganiswa kwedatha se-Fabric sincike ku-Data Factory kanye ne-Dataflow Gen2, ethola imisebenzi yoku... Ukuhlelwa kuzoba okuhlakaniphile, okuzenzakalelayo kakhudlwana, futhi kube nokufuduka okulula. kusuka kumapulatifomu akhona.

I-Dataflow Gen2: ukusebenza, ama-API omphakathi, kanye nokuxilongwa

Ku-Dataflow Gen2 sithola izici eziningana ekubukeni kuqala:

• Ukuhlela okuthuthukisiwe kwemibuzo eqondiwe ukulungisa i-logic endaweni oya kuyo ngqo kusuka endaweni yokubhala uqobo.
• I-computing ehlukanisiweokuvumela izingxenye zokugeleza kwedatha ukuthi zisebenze ngesikhathi esifanayo, kunciphisa isikhathi sokuhlola esiphelele.
• Landa ukuxilongwa ezingeni lokwenziwa, ngamaphakheji elogi ahlelekile okuhlaziya ukusebenza kanye nokuxazulula izigameko.
• Ama-API asesidlangalaleni ngokudala, ukubuyekeza, ukususa, ukuhlela, kanye nokuqapha ukugeleza kwedatha ngokuhlelekile.
• Amapharamitha omphakathi anokusekelwa kwe-CI/CDokuvumela ukuvuselela ukugeleza kwedatha ngokudlulisa amanani avela kumapayipi noma eminye imithombo.
• Idatha yakamuva ukuze uthole ukufinyelela okusheshayo ezintweni ezisetshenziswe muva nje ku-Power Query ribbon kanye naku-Modern Get Data.

Konke lokhu kuhambisana namakhono e- Hlola umbuzo wamandla ngokohlelo nge-RESTLokhu kuvula umnyango wokusebenzisa izikripthi ze-M njengengxenye yezinqubo ezenzakalelayo, ukuzihlanganisa ne-Spark, amapayipi, noma amathuluzi angaphandle, kusetshenziswa izixhumi ze-Power Query.

I-Data Factory: ukusebenza okuguquguqukayo, ukuxhumana kanye ne-dbt

Engxenyeni "ejwayelekile" yokuhlanganiswa, i-Data Factory ngaphakathi kwe-Fabric yethula:

• Ukulungiswa kokusebenza okuguquguqukayo ngomsebenzi wokukopisha, olungisa ngobuhlakani amapharamitha okusebenza ngokuya ngomongo wokucushwa kanye nokusebenza.
• Usekelo Lokushintsha Ukuthwebula Idatha (CDC) emsebenzini Wokukopisha, ukuphinda izinguquko kuphela (ukufaka, ukubuyekeza, ukususa) ngokuqhubekayo.
• Amasango asendaweni anenketho yokuthuthukisa ngesandla iphethwe kusukela ku-Fabric portal, i-API noma izikripthi.
• uxhumano lwakamuva, enezela izakhiwo zokugcina ezisetshenziswa ekuxhumaneni ukuze kube lula ukuhlolwa kwezimali kanye nokuphathwa komjikelezo wokuphila.
• Umsebenzi we-dbt womdabu, okuvumela ukuqhuba amaphrojekthi e-dbt ngaphakathi kwe-Fabric ngokuhlelwa okuhlanganisiwe, ukuhlolwa, ukubhalwa kwemibhalo kanye nokuphathwa.
• Yenza umsebenzi wephakheji ye-SSIS kumapayipi, ukusebenzisa amaphakheji e-SSIS kusukela ekuhleleni ngokwako ku-Fabric.

Ulwazi lomsebenzisi luphinde luthuthukiswe nge- Isikhethi sesayithi le-SharePoint (SharePoint Site Picker) egwema ukuthayipha ama-URL ngesandla, kanye nokusekelwa kwe-MCP kwe-Data Factory, ukuze abasizi be-AI bakwazi ukudala nokusabalalisa i-Dataflow Gen2 ngokumane nje besebenzisa imiyalelo yolimi lwemvelo.

Amathuluzi okufuduka kwedatha kanye nokuphindaphinda

I-Microsoft ikhuthaza kakhulu ukufuduka ku-Fabric ngamathuluzi amaningana athile:

• Ukuhlolwa Kokufuduka Kwendwangu Kwemboni Yedatha, ehlaziya ukulungiswa kwamapayipi e-ADF bese ithuthela lawo asekelwe endaweni yokusebenza yendwangu enemephu yokuxhumeka.
• Umsizi Wokufuduka We-Data Warehousemanje esingaxhuma ngqo endaweni yokugcina impahla ukuze siyithuthele e-Fabric Data Warehouse.
• Umsizi Wokufuduka wesizindalwazi se-SQL, okuhloswe ngayo ukufuduka kwemithwalo yemisebenzi ye-SQL Server endaweni, ngokungenisa i-schema nge-DACPAC, ukutholwa kokungahambelani kanye nezincomo.

Ngokuphathelene nokukopisha, kunikezwa ukwesekwa ukulingisa imithombo eminingi yokusebenza (Idathabheyisi ye-Azure ye-MySQL, i-Google BigQuery, i-SQL Server, njll.) ku-Fabric, enekhono lokulawula ukuthi yimaphi amathebula aphindaphindwayo, qala kabusha izinqubo zokulingisa nge-REST futhi, esimweni se-Databricks, faka izinqubomgomo ze-Unity Catalog ku-OneLake security.

Kufakwe futhi isixhumi sokuphindaphinda esivela eLakehouse esisebenzisa I-Delta Shintsha Idatha Yokuphakelwa, eveza izinguquko kumabhodi eLakehouse Delta eziya ezindaweni ezihambisanayo ngaphandle kokudinga ukuvuselela isondo ngezixazululo zasekhaya ze-CDC.

Ukuphepha, ukubusa okuthuthukisiwe, kanye nokuqapha ku-Fabric

Enye yezinto ezikhathazayo kakhulu kunoma iyiphi ipulatifomu yokuhlaziya ukuthi kanjani vikela idatha, lawula ukusetshenziswa, futhi uqaphe ukusetshenziswa kwezinsizaIndwangu ivuthwa ngokushesha kulezi zindawo.

Ukuphepha nokuvikelwa kwedatha e-OneLake

I-OneLake ingeza imodeli ephelele ye ukuphepha kokufinyelela idatha no:

• Izindima zokufinyelela idatha ze-lakehouse ezinezimvume ezilungisekayo ezivela kusixhumi esibonakalayo sokuphepha esisekelwe kufolda.
• Ukusekelwa kokuphepha kwezinqamuleli ukuze abantu besithathu bakwazi ukuhlonipha izinqubomgomo ezichaziwe.
• I-API yokuphepha kokufinyelela idatha ye-OneLake, evumela ukuphathwa kwemvume okuzenzakalelayo.
• Ukwelulwa kwemodeli kuzinjini zangaphandle (imvume yokuphepha ye-OneLake kubantu besithathu).

Ngesikhathi esifanayo, ukuvikelwa kuyandiswa nge Ukufinyelela okukhawulelwe kwe-DLP kuyo yonke idatha ehlelekile ku-OneLake (SQL, KQL, izindawo zokugcina impahla) futhi yethulwa I-DSPM ye-AI yama-Copilot endwangu kanye nama-ejenti edatha, eqapha ukusebenzisana kwe-AI ngolwazi olubucayi kanye nokuziphatha okuyingozi, ngokuhlanganiswa ne-Purview Audit kanye ne-eDiscovery.

Ngokuphathelene nobunikazi, izici ezifana nalezi ezilandelayo ziyavela: ubunikazi obuhlobene nezinto (isibonelo, iLakehouse kanye ne-Eventstream) ngama-REST API, aqeda ukuncika komnikazi emisebenzini ethile, kanye nokuqinisekiswa kwezinqamuleli ze-OneDrive kanye ne-SharePoint kusetshenziswa ubunikazi bendawo yokusebenza noma izinhloko zesevisi.

Ukubusa okuhlanganisiwe kanye nekhathalogi ye-OneLake

Okuhlangenwe nakho kokuphathwa kwedatha kuqiniswa yi- iphaneli entsha ephakathi kukhathalogi ye-OneLakelapho abanikazi bedatha bengabona khona umbono ohlanganisiwe wezinto abazidalile, bathole izincomo zesenzo sokubusa, futhi bafinyelele kuwo wonke amathuluzi atholakalayo okuthuthukisa ukuphepha nokuthobela imithetho.

Ngaphezu kwalokho, i- I-API Yokusesha Ikhathalogi ye-OneLake kanye nethuluzi le-MCP, elivumela ukuthola izinto kuyo yonke indawo ye-Fabric kusuka kuma-ejenti ekhodi noma e-AI, ngocingo olulodwa, kuhlonishwa izimvume zekhathalogi kanye nemethadatha.

Ukuqapha amandla, ukusetshenziswa kanye nomsebenzi

Indwangu inikeza izendlalelo eziningana zokubonakala:

• Ukuqapha indawo yokusebenza, okudala isizindalwazi ku-Fabric lapho amalogi kanye nama-metric avela ezintweni eziningi kuhlanganiswa khona (kufaka phakathi imisebenzi ye-Copy enokuqapha okuningiliziwe).
• Ukuqapha indawo yokusebenza yomsebenzi we-Copyngezilinganiso ezifana nomthamo, ivolumu yedatha, amakhodi amaphutha kanye nezikhathi, konke kuqondiswe ekuhlaziyweni okuhlanganisiwe.
• Umlando Wento ekusetshenzisweni kwezilinganiso zamandla, ngokubukwa kwezinsuku ezingu-30 kokusetshenziswa kwe-CU kwento ngayinye, okungahlungwa ngokwendawo yokusebenza kanye nohlobo.
• Ukuvikelwa kokuphakama kwezinga lokusebenzaokuvumela ukusetha imikhawulo yokusetshenziswa kwendawo yokusebenza ngayinye efasiteleni elihamba amahora angama-24, ukuvimba ngokuzenzakalelayo lezo ezidlulayo, nokumaka izindawo zokusebenza "njengezibalulekile" ukuze zingafakwa emikhawulweni.

Okuhambisana nalokhu, isethi yokuqala ye Ama-API Okuphatha Indwangu igxile ekutholeni izindawo zokusebenza, izinto kanye nemininingwane yokufinyelela komsebenzisi, ukwenza kube lula ukuqoqwa kwezinto ezisetshenziswayo kanye nokulawula ukufinyelela ngezikhathi ezithile.

Ukumodela ibhizinisi, ukuhlela, kanye nemisebenzi emisha

Ngaphandle kwesendlalelo sobuchwepheshe, iMicrosoft iyethula imithwalo yemisebenzi emisha egxile ebhizinisini mayelana ne-Fabric. Enye yezindlela ezivelele kakhulu yi-Fabric IQ, efuna ukuhlanganisa izincazelo zebhizinisi, idatha, kanye namamodeli amanxusa ahlakaniphile enza izinqumo ngokusekelwe embonweni ophelele wenhlangano.

Ngaphakathi kwe-Fabric IQ uzothola:

• I-Ontology (ukubuka kuqala), uhlobo lwento lapho izinhlangano, ubudlelwano, izakhiwo kanye nemingcele kuchazwa khona ngokolimi lwebhizinisi lenkampani.
• Uhlelo (ukuhlola kuqala), ipulatifomu engenakhodi yokuhlela, ukubika, ukuhlaziya, ukuhlanganisa, kanye nokuphatha ngokubambisana.

Ubuhlakani Besikhathi Sangempela nabo buvela umakhi wamawele wedijithali, into ekhethekile ekubumbeni amawele edijithali ngokusekelwe kudatha yesikhathi sangempela, ngenhloso yokwenza ngcono ukusebenza ngokomzimba, ukuqapha izimo kanye nokulingisa izimo.

Ngakolunye uhlangothi, kwethulwa Umsebenzi we-IQ yendwangu njengomthwalo womsebenzi ohlukile, kanye namathuluzi okusekela okubusa kanye nokuqondanisa amagama ayaqhubeka nokukhula, okuvala umbuthano phakathi kwamamodeli wedatha, i-logic yebhizinisi kanye nezinhlelo zokusebenza ze-AI/analytics.

Ukusebenza, ulwazi lomsebenzisi, kanye nokuthuthukiswa kokukhiqiza

Ukuze siphethe lokhu kubuyekezwa, kufanelekile ukugqamisa intuthuko eminingana ehlanganisa konke Azihlali zisematheni, kodwa zithonya kakhulu impilo yansuku zonke. wamaqembu.

Esigabeni se-Spark kanye ne-distributed computing, i-Fabric yethula:

• Isikhathi Sokusebenza Sendwangu 2.0 (ukubuka kuqala) nge-Apache Spark 4.0, i-Delta Lake 4.0, i-Java 21, i-Scala 2.13 kanye ne-Python 3.12 ku-Azure Linux 3.0.
• Ithuluzi lokuqhathanisa uhlelo lokusebenza lwe-Sparkokukuvumela ukuthi ukhethe futhi uqhathanise ukusebenza kwe-Spark okufika kwezine ngesikhathi esisodwa.
• I-Spark Diagnostic Emitter, eqoqa amalogi, amamethrikhi, kanye nemicimbi evela kuzinhlelo zokusebenza ze-Spark bese izithumela ezindaweni ezifana ne-Event Hubs, isitoreji, noma i-Log Analytics.
• Umtapo wolwazi lokuxilonga i-JobInsight, umtapo wolwazi wokuhlaziya ukwenziwa kwe-Spark okuqediwe ngama-API (imibuzo, imisebenzi, izigaba, imisebenzi, abaphathi, amalogi emicimbi).

Esigabeni sokugcina impahla, okulandelayo kuyangezwa: ukuhlanganiswa kwedatha Ukuze kuthuthukiswe ukusebenza nokunciphisa izindleko zokufinyelela, amakholomu e-IDENTITY okhiye abangesibo abakho, kanye nokulawulwa kwenguqulo kanye nokusekelwa kwe-CI/CD ngamaphrojekthi e-SQL Database ku-VS Code (ukulawulwa komthombo we-Warehouse).

Okuhlangenwe nakho komsebenzisi kwephothali ye-Fabric nakho kuyashintsha nge Ukuphequlula okunamathebhu kanye nokuhlola izintoLokhu kukuvumela ukuthi uvule izinto eziningi ngesikhathi esisodwa futhi ushintshe phakathi kwazo ngokushesha. Lokhu, kuhlanganiswe nentuthuko efana nokubopha okuzenzakalelayo kweLakehouse ku-Git kanye nesethi yezinsiza zokunakekelwa kweLakehouse (imisebenzi yokulungisa kanye nokuvuselela i-SQL endpoint), kunegalelo epulatifomu elula ukuyisebenzisa futhi esheshayo.

Okokugcina, izici ezifana Ukungeniswa/ukuthunyelwa kwencazelo yezinto ngobuningi (kokufuduka, amathempulethi kanye nokusekela ngokulondoloza imethadatha), i-REST yamafolda, ukwesekwa kwamapharamitha ekusebenzeni kwezinto kusuka ku-Activator, kanye nokulayisha idatha ye-OneLake ku-Excel ngekhathalogi ehlanganisiwe, kuqedela uhlelo lwe-ecosystem oluqala ukumboza cishe zonke izidingo ezivamile zethimba ledatha lesimanje.

Ngalo lonke leli qoqo lamakhono—kusukela ekubuseni okuhlanganisiwe, ukuphepha okuhlanganisiwe, kanye nokuhlelwa okuhlakaniphile, kuya ku-AI efakwe ku-SQL, ukuhlaziya kwesikhathi sangempela, amawele edijithali, kanye nama-ejenti e-MCP—i-Microsoft Fabric iqinisa isikhundla sayo njengeplatifomu yedatha ephelele lapho Isihluthulelo akusekho nje ukugcina nokubona idatha ngeso lengqondo, kodwa ukubusa, ukwenza ngokuzenzakalelayo, nokusebenzisa i-AI ukulawula yonke ingxenye yomjikelezo wokuphila kwedatha.okuvumela izinhlangano ukuthi zithuthukise izakhiwo zazo kancane kancane, zithuthe lokho esezinakho futhi zivumele izixazululo ezintsha ngokushesha kakhulu kunezindlela zendabuko.