Poyembekezera kuyamba kwa otaya lotsatira pa mlingo Takonzekera zolemba zazing'ono za mlembi ndi malangizo ofunikira popanga database. Tikukhulupirira kuti nkhaniyi ikuthandizani.
Zosungirako zili paliponse: kuyambira mabulogu osavuta kwambiri ndi zolemba mpaka machitidwe odalirika azidziwitso ndi malo ochezera akulu. Kaya nkhokweyo ndi yosavuta kapena yovuta sizofunika kwambiri chifukwa ndikofunika kuipanga moyenera. Pamene database idapangidwa mopanda kulingalira komanso popanda kumvetsetsa bwino cholinga chake, sizongogwira ntchito, koma ntchito yowonjezera ndi database idzakhala chizunzo chenicheni, nkhalango yosatheka kwa ogwiritsa ntchito. Nawa maupangiri opangira database omwe angakuthandizeni kupanga chinthu chothandiza komanso chosavuta kugwiritsa ntchito.
1. Dziwani kuti tebulo ndi chiyani komanso kapangidwe kake
Masiku ano, njira zachitukuko monga Scrum kapena RAD (Rapid Application Development) zimathandiza magulu a IT kupanga nkhokwe mwachangu. Komabe, mu kufunafuna nthawi, chiyeso ndi chachikulu kwambiri kuti tilowe pansi molunjika kumanga maziko, momveka bwino kulingalira chomwe cholingacho chiri, zomwe zotsatira zomaliza ziyenera kukhala.
Zili ngati kuti gululi likuyang'ana ntchito yogwira mtima, yofulumira, koma izi ndizosadabwitsa. Mukalowa mozama komanso mwachangu, m'pamenenso zimatenga nthawi kuti muzindikire ndikusintha zolakwika pamapangidwe a database.
Chifukwa chake chinthu choyamba chomwe muyenera kusankha ndikutanthauzira cholinga cha database yanu. Kodi nkhokweyo ikupangidwira ntchito yamtundu wanji? Kodi wogwiritsa ntchitoyo adzangogwira ntchito ndi zolemba zokha ndipo ayenera kumvetsera zochitika, kapena ali ndi chidwi kwambiri ndi kusanthula deta? Kodi maziko akhazikitsidwe kuti? Kodi idzatsata machitidwe a kasitomala kapena kungoyang'anira ubale wamakasitomala?
Mwamsanga gulu lopanga liyankha mafunso awa, njira yopangira ma database idzakhala yabwino.
2. Ndi deta yanji yomwe ndiyenera kusankha yosungira?
Konzekeranitu. Lingaliro la zomwe tsamba kapena dongosolo lomwe nkhokweyo ikupangidwira idzachita m'tsogolomu. Ndikofunikira kupitilira zofunikira zosavuta zaukadaulo. Chonde musayambe kuganizira za mitundu yonse ya data yomwe wogwiritsa ntchito angasunge. M'malo mwake, ganizirani ngati ogwiritsa ntchito adzatha kulemba zolemba, kukweza zikalata kapena zithunzi, kapena kusinthana mauthenga. Ngati ndi choncho, ndiye kuti muyenera kugawa malo awo mu database.
Gwirani ntchito ndi gulu, dipatimenti, kapena bungwe lomwe maziko ake adzathandizidwa mtsogolo. Lumikizanani ndi anthu amisinkhu yosiyanasiyana, kuchokera kwa akatswiri osamalira makasitomala mpaka atsogoleri a dipatimenti. Mwanjira iyi, mothandizidwa ndi mayankho, mupeza lingaliro lomveka bwino la zomwe kampaniyo ikufuna.
Mosapeweka, zosowa za ogwiritsa ntchito mkati mwa dipatimenti yomweyo zidzasemphana. Ngati mukukumana ndi izi, musaope kudalira zomwe mwakumana nazo ndikupeza mgwirizano womwe umagwirizana ndi maphwando onse ndikukwaniritsa cholinga chachikulu cha database. Khalani otsimikiza: m'tsogolomu mudzalandira +100500 mu karma ndi phiri la makeke.
3. Deta yachitsanzo mosamala
Pali mfundo zingapo zofunika kuziganizira potengera deta. Monga tanena kale, cholinga cha database chimasankha njira zomwe zingagwiritsidwe ntchito pojambula. Ngati tikupanga nkhokwe yoti tigwiritse ntchito pa intaneti (OLTP), mwa kuyankhula kwina popanga, kusintha ndi kufufuta ma rekodi, timagwiritsa ntchito mawonekedwe a transaction. Ngati database iyenera kukhala yogwirizana, ndiye kuti ndibwino kugwiritsa ntchito ma multidimensional modelling.
Pachitsanzo, ma data amalingaliro (CDM), thupi (PDM), ndi zomveka (LDM) amapangidwa.
Zitsanzo zamaganizidwe zimalongosola mabungwe ndi mitundu ya deta yomwe imaphatikizapo, komanso maubwenzi pakati pawo. Gawani deta yanu mumagulu omveka - zimapangitsa moyo kukhala wosavuta.
Chinthu chachikulu ndi kudziletsa, musapitirire.
Ngati bungwe liri lovuta kuliyika m'mawu amodzi kapena mawu amodzi, ndiye nthawi yoti mugwiritse ntchito ma subtypes (magulu a ana).
Ngati bungwe limakhala ndi moyo wake, liri ndi makhalidwe omwe amafotokoza khalidwe lake ndi maonekedwe ake, komanso maubwenzi ndi zinthu zina, ndiye kuti simungagwiritse ntchito subtype, komanso supertype (makolo).
Ngati munyalanyaza lamuloli, opanga ena adzasokonezeka mu chitsanzo chanu ndipo sangamvetse bwino deta ndi malamulo a momwe angasonkhanitsire.
Zitsanzo zamaganizidwe zimakhazikitsidwa pogwiritsa ntchito zomveka. Mitundu iyi ili ngati mapu apamsewu wamapangidwe ankhokwe. Muchitsanzo chomveka, mabungwe a data amazindikiridwa, mitundu ya data imatsimikiziridwa, ndipo mawonekedwe a fungulo lalamulo amatsimikiziridwa kuti amalamulira maubwenzi pakati pa data.
Kenako Logical Data Model ikufaniziridwa ndi nsanja yosankhidwa kale ya DBMS (database management system) ndipo Physical Model imapezedwa. Imalongosola momwe deta imasungidwa mwakuthupi.
4. Gwiritsani ntchito mitundu yoyenera ya data
Kugwiritsa ntchito mtundu wolakwika wa data kumatha kupangitsa kuti musakhale ndi zolondola kwambiri, zovuta pakujowina matebulo, zovuta kulumikiza mawonekedwe, ndi kukula kwamafayilo otupa.
Kuti zitsimikizire kukhulupirika kwa chidziwitso, chikhumbocho chiyenera kukhala ndi mitundu yokha ya deta yomwe imavomerezedwa. Ngati zaka zayikidwa munkhokwe, onetsetsani kuti gawolo likusunga manambala osapitilira 3.
Pangani mizati yochepa yopanda kanthu ndi mtengo wa NULL. Ngati mupanga mizati yonse ngati NULL, uku ndikulakwitsa kwakukulu. Ngati mukusowa ndime yopanda kanthu kuti mugwire ntchito inayake yamalonda, pamene deta sichidziwika kapena sichimveka bwino, ndiye kuti muzimasuka kuti mupange. Kupatula apo, sitingathe kudzaza mizati "Tsiku la imfa" kapena "Tsiku lothamangitsidwa" pasadakhale; sitiri olosera kuloza zala zathu kumwamba :-).
Mapulogalamu ambiri achitsanzo (ER/Studio, MySQL Workbench, SQL DBM, gliffy.com) data imakulolani kuti mupange ma prototypes a zigawo za data. Izi sizimangotsimikizira mtundu wolondola wa data, malingaliro ogwiritsira ntchito, ndi magwiridwe antchito abwino, komanso kuti mtengo wake ndiwofunika.
5. Pitani mwachilengedwe
Posankha ndime patebulo yoti mugwiritse ntchito ngati kiyi, nthawi zonse ganizirani magawo omwe wogwiritsa ntchito angasinthire. Osawasankha ngati kiyi - lingaliro loyipa. Chilichonse chikhoza kuchitika, koma muyenera kuwonetsetsa kuti ndi chapadera.
Ndikwabwino kugwiritsa ntchito kiyi yachilengedwe, kapena bizinesi. Lili ndi tanthauzo la semantic, kotero mudzapewa kubwereza mu database.
Pokhapokha ngati kiyi ya bizinesi ili yapadera (dzina loyamba, dzina lomaliza, malo) ndipo imabwerezedwa m'mizere yosiyana ya tebulo kapena iyenera kusintha, ndiye kuti kiyi yopangidwira iyenera kusankhidwa kukhala chinsinsi chachikulu.
6. Sangalalani mwachikatikati
Kuti mukonzekere bwino deta mu database, muyenera kutsatira malangizo angapo ndikusinthiratu nkhokwe. Pali mafomu asanu oyenera kutsatira.
Ndi normalization, mumapewa redundancy ndikuwonetsetsa kukhulupirika kwa zomwe zimagwiritsidwa ntchito patsamba lanu kapena tsamba lanu.
Monga nthawi zonse, zonse ziyenera kukhala zolimbitsa thupi, ngakhale normalization. Ngati pali matebulo ochulukirapo m'dawunilodi okhala ndi makiyi apadera omwewo, ndiye kuti mwatengeka ndikuwongolera nkhokwe. Kukhazikika kopitilira muyeso kumasokoneza magwiridwe antchito a database.
7. Yesani msanga, yesani pafupipafupi
Dongosolo loyesa ndi kuyesa koyenera ziyenera kukhala gawo la kapangidwe ka database.
Njira yabwino yoyesera deta yanu ndi kudzera mu Continuous Integration. Tsanzirani zochitika za "tsiku lachidziwitso" ndikuwona ngati milandu yonse yam'mphepete imayendetsedwa komanso momwe angagwiritsire ntchito ogwiritsa ntchito. Mwamsanga mutapeza nsikidzi, mudzapulumutsa nthawi ndi ndalama.
Awa ndi maupangiri asanu ndi awiri okha omwe mungagwiritse ntchito kuti mupange nkhokwe yopambana komanso yogwira ntchito bwino. Mukawatsatira, mudzapewa mutu wambiri m'tsogolomu. Malangizo awa ndi nsonga chabe ya iceberg mu database modelling. Pali ma hacks ambiri amoyo. Mumagwiritsa ntchito ziti?
Source: www.habr.com