Zvinoita sekunge iyo ndima yekushambadzira yepamhepo inofanirwa kunge iri yepamusoro tekinoroji uye otomatiki sezvinobvira. Ehe, nekuti hofori dzakadaro uye nyanzvi mumunda wavo seYandex, Mail.Ru, Google uye Facebook vanoshanda ipapo. Asi, sezvazvakazoitika, hapana muganho wekukwana uye pane nguva dzose chimwe chinhu chekugadzirisa.
Boka rekukurukurirana Dentsu Aegis Network Russia ndiye mutambi mukuru mumusika wekushambadza kwedhijitari uye arikushingairira kudyara muhunyanzvi, achiedza kukwidziridza nekuita otomatiki maitiro ayo ebhizinesi. Rimwe rematambudziko asina kugadziriswa emusika wekushambadza wepamhepo ibasa rekuunganidza zviverengero pamishandirapamwe yekushambadzira kubva kumapuratifomu eInternet akasiyana. Mhinduro yedambudziko iri pakupedzisira yakaguma nekugadzirwa kwechigadzirwa D1.Digital (verenga seDiVan), budiriro yatinoda kutaura nezvayo.
Sei?
1. Panguva yekutanga kwepurojekiti, pakanga pasina kana chigadzirwa chimwe chete chakagadzirirwa pamusika chakagadzirisa dambudziko rekugadzirisa kuunganidza kwehuwandu pamashambadziro ekushambadzira. Izvi zvinoreva kuti hapana mumwe kunze kwedu isu achazadzisa zvatinoda.
Masevhisi akadai seImprovado, Roistat, Supermetrics, SegmentStream inopa kubatanidzwa nemapuratifomu, masocial network uye Google Analitycs, uye zvakare inoita kuti zvikwanise kuvaka madhibhodhi ekuongorora kuti aongorore zviri nyore uye kutonga kwemashambadziro ekushambadzira. Tisati tatanga kugadzira chigadzirwa chedu, takaedza kushandisa mamwe masisitimu aya kuunganidza data kubva kumasaiti, asi, zvinosuruvarisa, havana kukwanisa kugadzirisa matambudziko edu.
Dambudziko guru nderekuti zvigadzirwa zvakaedzwa zvakavakirwa padhata masosi, zvichiratidza nhamba dzekuiswa nesaiti, uye hazvina kupa kugona kuunganidza zviverengero pamishandirapamwe yekushambadzira. Iyi nzira haina kutibvumira kuona nhamba kubva kunzvimbo dzakasiyana-siyana munzvimbo imwe uye kuongorora mamiriro emushandirapamwe wose.
Chimwe chikonzero ndechekuti pamatanho ekutanga zvigadzirwa zvaive zvakanangana nemusika wekuMadokero uye hazvina kutsigira kubatanidzwa nenzvimbo dzeRussia. Uye kune idzo saiti dzakabatanidzwa, ese ma metrics aidiwa aisagara achitorwa aine ruzivo rwakakwana, uye kusanganisa kwaisagara kuri nyore uye kwakajeka, kunyanya pazvaidiwa kuwana chimwe chinhu chisiri muhurongwa hwehurongwa.
Kazhinji, isu takasarudza kusajaira kune yechitatu-bato zvigadzirwa, asi takatanga kugadzira zvedu ...
2. Musika wekushambadzira wepamhepo uri kukura gore negore, uye muna 2018, maererano nemabhajeti ekushambadzira, yakakunda musika mukuru wekushambadzira weTV. Saka pane chikero.
3. Kusiyana nemusika wekushambadza weTV, uko kutengeswa kwekushambadzira kwekutengesa kunoitwa monopol, kune vazhinji varidzi vega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega ega. Sezvo mushandirapamwe wekushambadzira, sekutonga, unomhanya panzvimbo dzinoverengeka kamwechete, kuti unzwisise mamiriro ekushambadzira, zvinodikanwa kuunganidza mishumo kubva kune ese mawebhusaiti uye kuisanganisa muhumwe mushumo mukuru unozoratidza mufananidzo wese. Izvi zvinoreva kuti pane mukana we optimization.
4. Zvaiita kwatiri kuti varidzi vezvokushambadza zvinyorwa paInternet vatova nezvivako zvekuunganidza zviverengero uye kuzviratidzira mumaakaundi ekushambadzira, uye vachakwanisa kupa API yeiyi data. Izvi zvinoreva kuti zvinogoneka nehunyanzvi kuishandisa. Ngatitii pakarepo kuti zvakava zvisiri nyore.
Kazhinji, zvese zvinodikanwa zvekuitwa kweprojekiti zvaive pachena kwatiri, uye isu takamhanya kuunza purojekiti kuhupenyu ...
Grand Plan
Kutanga, takagadzira chiono cheyakanaka system:
- Mishandirapamwe yekushambadzira kubva ku1C corporate system inofanirwa kuiswa mairi nemazita avo, nguva, mabhajeti uye kuiswa pamapuratifomu akasiyana.
- Pakuiswa kwega kwega mukati memushandirapamwe wekushambadzira, zvese zvinogoneka zviverengero zvinofanirwa kudhaunirodwa kubva kunzvimbo dzirikuiswa, senge nhamba yezviratidziro, kudzvanya, maonero, nezvimwe.
- Mimwe mishandirapamwe yekushambadzira inoteedzerwa uchishandisa yechitatu-bato yekutarisa neinonzi adserving masisitimu akadai saAdriver, Weborama, DCM, nezvimwe. Kune zvakare maindasitiri eInternet mita muRussia - iyo Mediascope kambani. Zvinoenderana nehurongwa hwedu, data kubva kune yakazvimirira uye maindasitiri ekutarisa anofanirwawo kuiswa otomatiki mumashambadziro anoenderana.
- Mishandirapamwe yakawanda yekushambadzira paInternet inonangana nezvimwe zviito zvinotarirwa (kutenga, kufona, kusaina bvunzo yekutyaira, nezvimwewo), izvo zvinotevedzwa uchishandisa Google Analytics, uye manhamba ayo akakoshawo pakunzwisisa mamiriro emushandirapamwe uye. inofanira kuiswa muchishandiso chedu .
Chinhu chekutanga chine lumpy
Tichifunga nezvekuzvipira kwedu kune inochinjika misimboti yekuvandudza software (yakareruka, zvinhu zvese), takafunga kutanga tagadzira MVP tobva taenda kune chinangwa chatakanga taronga zvakare.
Takasarudza kuvaka MVP zvichienderana nechigadzirwa chedu DANBo (Dentsu Aegis Network Board), inova yewebhu application ine ruzivo rwakakwana pamakasitoma ekushambadzira evatengi vedu.
KuMVP, chirongwa chakarerutswa zvakanyanya sezvinobvira maererano nekuita. Isu takasarudza runyoro rushoma rwemapuratifomu ekubatanidza. Aya ndiwo aive mapuratifomu makuru, akadai seYandex.Direct, Yandex.Display, RB.Mail, MyTarget, Adwords, DBM, VK, FB, uye main adserving systems Adriver uye Weborama.
Kuti uwane nhamba pamasaiti kuburikidza neAPI, takashandisa account imwe chete. Maneja weboka revatengi aida kushandisa otomatiki kuunganidzwa kwenhamba pamushandirapamwe wekushambadzira aifanira kutanga agovera mukana kune anodiwa ekushambadzira mishandirapamwe pasaiti kuaccount account.
Inotevera ndeye mushandisi wesystem DANBo aifanira kurodha faira reimwe fomati muExcel system, yaive neruzivo rwese nezvekuiswa (kushambadzira mushandirapamwe, chikuva, fomati, nguva yekuisa, zviratidzo zvakarongwa, bhajeti, nezvimwewo) uye zviziviso zveanowirirana ekushambadzira kushambadzira pa nzvimbo nemakaunda mune adserving systems.
Zvairatidza, kutaura chokwadi, zvinotyisa:
Iyo data yakatorwa yakachengetwa mudhatabhesi, uyezve masevhisi akaparadzana akaunganidza zviziviso zvemushandirapamwe pamasaiti kubva kwavari uye kudhawunirodha nhamba pazviri.
Kune yega yega saiti, yakaparadzana windows sevhisi yakanyorwa, iyo kamwe pazuva yaienda pasi peakaundi imwe yebasa mune saiti API uye kudhawunirodha manhamba eakatsanangurwa maID emushandirapamwe. Chinhu chimwe chete chakaitika neadserving systems.
Iyo data yakatorwa yakaratidzwa pane iyo interface muchimiro chediki tsika dashboard:
Tisingazvitarisire isu, MVP yakatanga kushanda uye yakatanga kudhawunirodha nhamba dzazvino pamishandirapamwe yekushambadzira paInternet. Isu takaisa sisitimu kune akati wandei vatengi, asi pakuedza kuyera, takasangana nematambudziko akakura:
- Dambudziko guru raive rakaomarara rekugadzirira data rekurodha musystem. Zvakare, iyo data yekuisa yaifanira kushandurwa kuita fomati yakasimba isati yarodha. Zvaive zvakakodzera kusanganisa zviziviso zvenhengo kubva kunzvimbo dzakasiyana mufaira rekurodha. Isu takatarisana nenyaya yekuti zvakaoma kune vashandisi vasina kudzidziswa kuti vatsanangure kwekuwana zviziviso izvi pasaiti uye kuti mufaira ravanoda kupinzwa kupi. Tichifunga nezvehuwandu hwevashandi mumadhipatimendi ari kuita mishandirapamwe panzvimbo nenzvimbo, izvi zvakakonzera rutsigiro rwakakura kudivi redu, izvo zvatakanga tisina kufara nazvo.
- Rimwe dambudziko nderekuti haasi ese mapuratifomu ekushambadzira aive nemaitiro ekugovera mukana wekushambadzira mishandirapamwe kune mamwe maakaundi. Asi kunyangwe dai nzira yekutuma yaivepo, havazi vese vashambadziri vakada kupa mukana kumaakaundi ebato ravo.
- Chinhu chakakosha kwaive kushatirwa kwakamutsa pakati pevashandisi nenyaya yekuti zvese zvakarongwa zviratidzo uye ruzivo rwekuisa izvo zvavanotopinda mune yedu 1C accounting system, ivo vanofanirwa kupinda mukati. DANBo.
Izvi zvakatipa pfungwa yekuti iyo yekutanga sosi yeruzivo nezve kuiswa kunofanirwa kuve yedu 1C sisitimu, iyo data yese inopinzwa nemazvo uye nenguva (iyo pfungwa apa ndeyekuti invoice inogadzirwa zvichibva pane 1C data, saka kwakaringana kupinda kwedata mu1C. chinhu chakakosha kune wese KPI). Aya ndiwo maitiro matsva ehurongwa hwehurongwa ...
Concept
Chinhu chekutanga chatakafunga kuita kwaive kuparadzanisa sisitimu yekuunganidza zviverengero zvekushambadzira mishandirapamwe paInternet kuita chimwe chigadzirwa - D1.Digital.
Mune pfungwa itsva, takasarudza kurodha mukati D1.Digital ruzivo rwekushambadzira mishandirapamwe uye kuiswa mukati mawo kubva ku1C, uye wobva wadhonza nhamba kubva kumasaiti uye AdServing masisitimu kune aya anoiswa. Izvi zvaifanirwa kurerutsa zvakanyanya hupenyu hwevashandisi (uye, semazuva ese, kuwedzera rimwe basa kune vanogadzira) uye kuderedza huwandu hwekutsigirwa.
Dambudziko rekutanga ratakasangana naro raive rehurongwa hwesangano uye raienderana nekuti hatina kuwana kiyi kana chiratidzo chataigona kufananidza masangano kubva kuhurongwa hwakasiyana nemishandirapamwe nekuiswa kubva ku1C. Icho chokwadi ndechekuti maitiro mukambani yedu akagadzirwa nenzira yekuti mishandirapamwe yekushambadzira inopinzwa mumasisitimu akasiyana nevanhu vakasiyana (media kuronga, kutenga, nezvimwewo).
Kuti tigadzirise dambudziko iri, taifanira kugadzira yakasarudzika hashi kiyi, DANBoID, yaizobatanidza masangano mune akasiyana masisitimu pamwe chete, uye iyo inogona kuzivikanwa zviri nyore uye yakasarudzika mumaseti edata akadhawunirodha. Ichi chiziviso chinogadzirwa mukati me1C system yekuiswa kwega kwega uye inoendeswa kumishandirapamwe, kuiswa uye macounter pamasaiti ese uye mune ese AdServing masisitimu. Kuita tsika yekuisa DANBoID munzvimbo dzese kwakatora nguva, asi isu takakwanisa kuzviita :)
Ipapo takaona kuti haasi ese mawebhusaiti ane API yekuunganidza otomatiki manhamba, uye kunyangwe ayo ane API, haadzose data rese rinodiwa.
Panguva ino, takasarudza kudzikisa zvakanyanya rondedzero yemapuratifomu ekubatanidza uye kutarisa pamapuratifomu makuru anobatanidzwa muhuwandu hwakawanda hwekushambadzira kushambadzira. Rondedzero iyi inosanganisira vese vatambi vakuru mumusika wekushambadzira (Google, Yandex, Mail.ru), masocial network (VK, Facebook, Twitter), makuru AdServing uye analytics masisitimu (DCM, Adriver, Weborama, Google Analytics) uye mamwe mapuratifomu.
Mazhinji emasaiti atasarudza aive neAPI yaipa metrics yataida. Muzviitiko apo pakanga pasina API kana yakanga isina dhata inodiwa, takashandisa mishumo inotumirwa zuva nezuva kuhofisi yedu email kurodha data (mune mamwe masisitimu zvinokwanisika kugadzirisa mishumo yakadai, mune dzimwe takabvumirana pakuvandudzwa kwemishumo yakadaro. zvedu).
Pakuongorora data kubva kune akasiyana saiti, takaona kuti hierarchy yemasangano haina kufanana mumasisitimu akasiyana. Uyezve, ruzivo runofanira kutorwa mune zvakasiyana zvakadzama kubva kune akasiyana masisitimu.
Kugadzirisa dambudziko iri, iyo SubDANBoID pfungwa yakagadzirwa. Pfungwa yeSubDANBoID iri nyore, tinomaka chikamu chikuru chemushandirapamwe pasaiti neiyo yakagadzirwa DANBoID, uye isu tinorodha ese akaiswa masangano ane akasiyana saiti anozivisa uye fomu SubDANBoID zvinoenderana neiyo DANBoID musimboti + identifier yedanho rekutanga. nested entity + identifier of the second level nested entity +... Nzira iyi yakatibvumira kubatanidza mishandirapamwe yekushambadzira mumasisitimu akasiyana uye kudhawunirodha huwandu hwakadzama pazviri.
Taifanirawo kugadzirisa dambudziko rekuwana mishandirapamwe pamapuratifomu akasiyana. Sezvatakanyora pamusoro, nzira yekugovera mukana kumushandirapamwe kune yakasiyana account account haishande nguva dzose. Naizvozvo, taifanira kugadzira zvivakwa zvekubvumidza otomatiki kuburikidza neOAuth tichishandisa tokens uye masisitimu ekugadzirisa aya tokeni.
Gare gare mune chinyorwa tichaedza kutsanangura zvakadzama mavakirwo emhinduro uye ruzivo rwehunyanzvi hwekuita.
Solution architecture 1.0
Patakatanga kushandiswa kwechigadzirwa chitsva, takanzwisisa kuti isu takabva tatoda kupa mukana wekubatanidza masaiti matsva, saka takasarudza kutevera nzira ye microservice architecture.
Kana tichigadzira dhizaini, takaparadzanisa zvinongedzo kune ese ekunze masisitimu - 1C, mapuratifomu ekushambadzira uye masisitimu ekushambadzira - mumasevhisi akasiyana.
Pfungwa huru ndeyekuti zvese zvinongedzo kumasaiti zvine API yakafanana uye maadapter anounza saiti API kune interface yakanakira isu.
Pakati pechigadzirwa chedu iwebhu application, inova monolith yakagadzirwa nenzira yekuti inogona kuparadzaniswa nyore nyore kuita masevhisi. Ichi chishandiso chine basa rekugadzirisa iyo data yakatorwa, kuunganidza zviverengero kubva kune akasiyana masisitimu uye kuzvipa kune vashandisi system.
Kuti titaurirane pakati pezvibatanidza newebhu application, taifanira kugadzira imwe sevhisi, yatakadaidza kuti Connector Proxy. Inoita mabasa eSevhisi Discovery uye Task Scheduler. Iyi sevhisi inoita mabasa ekuunganidza data kune yega yega yekubatanidza manheru ega ega. Kunyora sevhisi layer kwaive nyore pane kubatanidza meseji broker, uye kwatiri zvaive zvakakosha kuti tiwane mhedzisiro nekukurumidza sezvinobvira.
Kuti zvive nyore uye nekumhanya kwekusimudzira, isu takasarudzawo kuti masevhisi ese ave Webhu API. Izvi zvakaita kuti zvikwanise kukurumidza kuunganidza chiratidzo-che-pfungwa uye kuona kuti dhizaini yese inoshanda.
Rimwe rakaparadzana, rakaomesesa basa raive kugadzira mukana wekuunganidza data kubva kumaakaundi akasiyana, ayo, sezvatakasarudza, anofanirwa kuitwa nevashandisi kuburikidza newebhu interface. Iyo ine matanho maviri akapatsanurwa: chekutanga, mushandisi anowedzera chiratidzo chekuwana iyo account kuburikidza neOAuth, uyezve anogadzirisa kuunganidzwa kwe data kumutengi kubva kune yakatarwa account. Kuwana chiratidzo kuburikidza neOAuth kwakakosha nekuti, sezvatakatonyora, hazvigoneke nguva dzose kupa mukana kune iyo yaunoda account pane saiti.
Kugadzira nzira yepasirese yekusarudza account kubva kumasaiti, taifanira kuwedzera nzira kune yekubatanidza API inodzosera JSON Schema, iyo inopihwa kuita fomu uchishandisa yakagadziridzwa JSONEditor chikamu. Nenzira iyi, vashandisi vakakwanisa kusarudza maakaundi kubva kwairi kutora data.
Kuti tienderane nemiganhu yekukumbira iripo pamasayiti, tinobatanidza zvikumbiro zvezvirongwa mukati mechiratidzo chimwe, asi tinogona kugadzirisa zviratidzo zvakasiyana zvakafanana.
Isu takasarudza MongoDB sechichengetedzo che data rakaremerwa kune ese ari maviri webhu application uye majoini, izvo zvakatitendera kuti tisanyanya kunetseka nezve data chimiro pamatanho ekutanga ebudiriro, kana chinhu chemhando yechishandiso ichichinja mamwe mazuva ese.
Nenguva isipi takaona kuti haisi yese data inokodzera muMongoDB uye, semuenzaniso, zviri nyore kuchengetedza nhamba dzezuva nezuva mudura rehukama. Naizvozvo, kune vanobatanidza vane chimiro chedata chakanyanya kukodzera dhatabhesi rehukama, takatanga kushandisa PostgreSQL kana MS SQL Server sekuchengetedza.
Izvo zvakasarudzwa zvivakwa uye matekinoroji zvakatibvumira kuvaka uye kutanga iyo D1.Digital chigadzirwa nekukurumidza. Kwemakore maviri ekugadzirwa kwechigadzirwa, takagadzira 23 zvinongedzo kumasaiti, takawana ruzivo rwakakosha kushanda nevechitatu-bato APIs, takadzidza kudzivirira misungo yemasaiti akasiyana, ayo ega ega aive neayo, akabatsira mukuvandudza maAPIs kune angangoita matatu masaiti. , yakatora otomatiki ruzivo pamusoro pemishandirapamwe inosvika zviuru gumi nezvishanu uye kune anopfuura 3 kuiswa, yakaunganidza yakawanda mhinduro kubva kune vashandisi pamusoro pekushanda kwechigadzirwa uye yakakwanisa kushandura nzira huru yechigadzirwa kakawanda, zvichienderana nemhinduro iyi.
Solution architecture 2.0
Makore maviri apfuura kubva pakatanga budiriro D1.Digital. Kuramba kuchiwedzera kwemutoro pane sisitimu uye kubuda kweakawanda uye akawanda matsva data masosi zvishoma nezvishoma kwakaratidza matambudziko mune iripo mhinduro yekuvaka.
Dambudziko rekutanga rine chekuita nehuwandu hwe data yakatorwa kubva kumasaiti. Isu takanga takatarisana nenyaya yekuti kuunganidza nekuvandudza data rese rinodiwa kubva kumasaiti makuru akatanga kutora nguva yakawandisa. Semuyenzaniso, kuunganidza data kubva kuAdRiver adserving system, yatinoronda nayo manhamba enzvimbo zhinji, zvinotora maawa gumi nemaviri.
Kugadzirisa dambudziko iri, takatanga kushandisa marudzi ese emishumo kudhawunirodha data kubva kumasaiti, tiri kuyedza kuvandudza API yavo pamwe nemasaiti kuitira kuti kumhanya kwekushanda kwayo kukwane zvatinoda, uye kufananidza kurodha data zvakanyanya sezvinobvira.
Rimwe dambudziko rine chekuita nekugadziriswa kwe data yakatorwa. Ikozvino, kana huwandu hutsva hwekuiswa hwasvika, nhanho dzakawanda dzekuverengera mametrics dzinotangwa, izvo zvinosanganisira kurodha data mbishi, kuverenga akaunganidzwa metrics yesaiti yega yega, kuenzanisa data kubva kwakasiyana kune imwe neimwe, uye kuverenga pfupiso metrics yemushandirapamwe. Izvi zvinokonzeresa mutoro wakawanda pawebhu application iyo inoita zvese kuverenga. Nguva dzinoverengeka, panguva yekudzokorora maitiro, iyo application yakapedza ndangariro dzese pane sevha, nezve 10-15 GB, iyo yakave nemhedzisiro inokuvadza pabasa revashandisi nehurongwa.
Matambudziko akaonekwa uye zvirongwa zvechimbichimbi zvekuwedzera kuvandudzwa kwechigadzirwa zvakatitungamira kune kudiwa kwekutarisa zvakare chivakwa chekushandisa.
Takatanga nema connectors.
Isu takaona kuti zvese zvinobatanidza zvinoshanda zvinoenderana neiyo modhi imwechete, saka isu takavaka pombi yekumisikidza iyo yekugadzira chinongedzo iwe waingofanirwa kuronga logic yematanho, mamwe aive pasi rose. Kana imwe yekubatanidza ichida kuvandudzwa, saka isu tinobva tangoiendesa kune nyowani nyowani panguva imwe chete iyo yekubatanidza iri kuvandudzwa.
Panguva imwecheteyo, takatanga kutumira zvinongedzo kuDocker uye Kubernetes.
Isu takaronga kutamira kuKubernetes kwenguva yakareba, tichiyedza CI/CD marongero, asi takatanga kufamba chete kana imwe yekubatanidza, nekuda kwekukanganisa, yakatanga kudya zvinopfuura makumi maviri GB yendangariro paseva, ichiuraya mamwe maitiro. . Munguva yekuferefeta, chinongedzo chakatamiswa kuenda kuKubernetes cluster, kwachakazosara, kunyangwe chikanganiso chagadziriswa.
Nekukurumidza takaona kuti Kubernetes yaive nyore, uye mukati memwedzi mitanhatu takatamisa zvinongedzo zvinomwe uye Connectors Proxy, iyo inoshandisa zvakanyanya zviwanikwa, kuboka rekugadzira.
Tichitevera zvibatanidza, takasarudza kushandura dhizaini yeimwe yese yekushandisa.
Dambudziko guru nderekuti data rinobva kune zvibatanidza kune proxies mumabhechi makuru, uye yobva yarova iyo DANBoID uye inotumirwa kune yepakati web application yekugadziriswa. Nekuda kwehuwandu hukuru hwekuverengera metrics, kune mutoro wakakura pane application.
Zvakaratidzawo kuti zvakaoma kutarisa mamiriro ega ega data rekuunganidza mabasa uye kushuma kukanganisa kunoitika mukati mezvibatanidza kune yepakati web application kuitira kuti vashandisi vaone zviri kuitika uye nei data risiri kuunganidzwa.
Kugadzirisa matambudziko aya, takagadzira zvivakwa 2.0.
Musiyano mukuru pakati peiyo vhezheni itsva yezvivakwa ndeyekuti panzvimbo yeWebhu API, isu tinoshandisa RabbitMQ neraibhurari yeMassTransit kuchinjanisa mameseji pakati pemasevhisi. Kuti tiite izvi, taifanira kunyora zvachose maConnectors Proxy, tichiiita Connectors Hub. Iro zita rakashandurwa nekuti iro basa guru resevhisi harisisiri mukutumira zvikumbiro kune vanobatanidza uye kumashure, asi mukutonga kuunganidzwa kwemametrics kubva kune zvinobatanidza.
Kubva pakati pewebhu application, takaparadzanisa ruzivo rwekuiswa uye nhamba kubva kune masaiti kuita masevhisi akaparadzana, izvo zvakaita kuti zvibvire kubvisa zvisingakoshi recalculation uye kuchengetedza chete yakaverengerwa uye yakaunganidzwa manhamba padanho rekuiswa. Isu takanyora zvakare uye nekugonesa iyo logic yekuverenga basic statistics zvichibva pane yakabikwa data.
Panguva imwecheteyo, isu tiri kutamisa ese masevhisi uye maapplication kuDocker neKubernetes kuita kuti mhinduro ive nyore kuyera uye nyore kubata.
Tirikupi manje
Uchapupu-hwe-pfungwa yekuvaka 2.0 chigadzirwa D1.Digital yakagadzirira uye inoshanda munzvimbo yekuyedza ine mashoma seti yezvikoni. Chasara kuita kunyorazve mamwe makumi maviri ekubatanidza papuratifomu nyowani, kuyedza kuti data rakaremerwa nemazvo uye ese metrics akaverengerwa nemazvo, uye kuburitsa dhizaini yese mukugadzira.
Muchokwadi, maitiro aya achaitika zvishoma nezvishoma uye isu tichafanirwa kusiya kumashure kuenderana nekare APIs kuchengetedza zvese kushanda.
Zvirongwa zvedu zvechimbichimbi zvinosanganisira kuvandudzwa kwezvibatanidza zvitsva, kubatanidzwa nemasisitimu matsva uye kuwedzera mamwe metrics kune seti yedata yakatorwa kubva kune akabatana masaiti uye adserving masisitimu.
Isu tinoronga zvakare kuendesa zvese zvikumbiro, kusanganisira iyo yepakati webhu application, kuDocker neKubernetes. Yakasanganiswa neiyo dhizaini nyowani, izvi zvinonyanya kurerutsa kuendesa, kutarisa uye kutonga kwezvinhu zvinodyiwa.
Imwe pfungwa ndeyekuyedza nesarudzo yedhatabhesi yekuchengetedza nhamba, iyo parizvino yakachengetwa muMongoDB. Isu takatotamisa akati wandei ekubatanidza kuSQL dhatabhesi, asi ipapo mutsauko unenge usingaonekwe, uye kune akaunganidzwa manhamba masikati, ayo anogona kukumbirwa kwenguva yekupokana, kuwana kwacho kunogona kunge kwakakomba.
Kazhinji, zvirongwa zvakakura, ngatienderere mberi :)
Vanyori vechinyorwa R&D Dentsu Aegis Network Russia: Georgy Ostapenko (), Mikhail Kotsik ()
Source: www.habr.com