{"id":34157,"date":"2019-10-31T21:56:43","date_gmt":"2019-10-31T18:56:43","guid":{"rendered":"https:\/\/prohoster.info\/blog\/kak-nachat-devops-transformatsiyu\/"},"modified":"2019-10-31T21:56:43","modified_gmt":"2019-10-31T18:56:43","slug":"kak-nachat-devops-transformatsiyu","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sq\/blog\/administrirovanie\/kak-nachat-devops-transformatsiyu","title":{"rendered":"Si t\u00eb filloni transformimin DevOps","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

N\u00ebse nuk e kuptoni se \u00e7far\u00eb \u00ebsht\u00eb DevOps, ja nj\u00eb p\u00ebrmbledhje e shkurt\u00ebr. DevOps \u00ebsht\u00eb nj\u00eb grup praktikash q\u00eb redukojn\u00eb frik\u00ebrat e inxhinier\u00ebve<\/strong> dhe reduktojn\u00eb numrin e d\u00ebshtimeve n\u00eb prodhimin e softuerit. N\u00eb p\u00ebrgjith\u00ebsi, ato gjithashtu redukojn\u00eb koh\u00ebn p\u00ebr t\u00eb hyr\u00eb n\u00eb treg<\/strong> \u2014 periudha nga ideja deri n\u00eb dor\u00ebzimin e produktit p\u00ebrfundimtar te klient\u00ebt, e cila lejon q\u00eb t\u00eb kryhen shpejt eksperimentet biznesore<\/strong>.<\/p>\n

Si t\u00eb filloni transformimin DevOps? N\u00ebse flasim shkurt, zgjidhni sh\u00ebrbimin me t\u00eb cilin do t\u00eb filloni procesin, identifikoni ata q\u00eb kan\u00eb lidhje me sh\u00ebrbimin, nd\u00ebrtoni Hart\u00ebn e Vler\u00ebs, krijoni nj\u00eb ekip p\u00ebrkoh\u00ebsisht q\u00eb do t\u00eb merret me transformimin p\u00ebr t\u00eb parin koh\u00eb dhe vendosni nj\u00eb detyr\u00eb p\u00ebr t\u00eb. P\u00ebrs\u00ebritni ciklin sa her\u00eb q\u00eb nevojitet.<\/p>\n

\"Si<\/p>\n

Nj\u00eb plan i detajuar p\u00ebr transformimin DevOps me shembuj dhe udh\u00ebzime \u00ebsht\u00eb n\u00eb shpjegim referat<\/a><\/noindex> Andrej Aleksandrov<\/b> \u2014 inxhinierit n\u00eb kompanin\u00eb Express42, e cila ofron k\u00ebshilla p\u00ebr zhvillimin e DevOps, duke e p\u00ebrshpejtuar k\u00ebt\u00eb proces, sepse tashm\u00eb ka nd\u00ebrtuar hart\u00ebn e pengesave. N\u00ebse ju duket se transformimi nuk \u00ebsht\u00eb p\u00ebr ju, ose keni nj\u00eb specifik\u00eb q\u00eb praktikat DevOps nuk p\u00ebrputhen, - p\u00ebrdorni raportin si nj\u00eb udh\u00ebzues p\u00ebr gjetjen dhe eliminimin e pengesave.
\n<\/a><\/noindex>
\nN\u00ebse jeni t\u00eb shqet\u00ebsuar p\u00ebr transformimin DevOps, at\u00ebher\u00eb keni nj\u00eb kompani t\u00eb madhe dhe duhet ta zgjeroni ngadal\u00eb k\u00ebt\u00eb proces n\u00eb t\u00ebr\u00eb struktur\u00ebn. Deri sa ka nevoj\u00eb p\u00ebr t\u00eb transformuar ekipin ose p\u00ebr t\u00eb eliminuar nj\u00eb penges\u00eb, algoritmi m\u00eb posht\u00eb mund t\u00eb p\u00ebrs\u00ebritet.<\/p>\n

Zgjedhja e sh\u00ebrbimit<\/h2>\n

\nPlani \u00ebsht\u00eb p\u00ebrgatitur, le t\u00eb fillojm\u00eb me hapin e par\u00eb - zgjedhjen e sh\u00ebrbimit. Kriti i par\u00eb - koha e jet\u00ebs<\/strong>: ka sh\u00ebrbime t\u00eb vjetra - legac\u00eb, dhe t\u00eb reja. Mund t\u00eb filloni nga t\u00eb dyja.<\/p>\n

T\u00eb zgjidhni nj\u00eb sh\u00ebrbim t\u00eb ri ka kuptim<\/strong>. Ai \u00ebsht\u00eb i ri, nuk ka ende nj\u00eb proces t\u00eb stabilizuar t\u00eb pun\u00ebs n\u00eb ekipin q\u00eb merret me t\u00eb. Nuk ka nj\u00eb mal t\u00eb borxhit teknik p\u00ebrreth, nuk \u00ebsht\u00eb e nevojshme ta riparoni at\u00eb vazhdimisht. Mund t\u00eb b\u00ebjm\u00eb me t\u00eb gjith\u00e7ka q\u00eb duam.<\/p>\n

N\u00eb rastin e nj\u00eb sh\u00ebrbimi t\u00eb vjet\u00ebr, ka probleme q\u00eb lidhen me t\u00eb ndryshimi gjithmon\u00eb \u00ebsht\u00eb i v\u00ebshtir\u00eb<\/strong>. Aty ka nj\u00eb grup kufizimesh serioze, por ndoshta jan\u00eb njer\u00ebz q\u00eb jan\u00eb t\u00eb gatsh\u00ebm t\u00eb presin gjith\u00e7ka - ata jan\u00eb lodhur dhe duan t\u00eb b\u00ebjn\u00eb di\u00e7ka ndryshe, sepse iu shkakton dhimbje.<\/p>\n

Puna me nj\u00eb sh\u00ebrbim t\u00eb vjet\u00ebr krijon nj\u00eb precedent t\u00eb fort\u00eb<\/strong> n\u00eb kompanin\u00eb tuaj - mund t\u00eb b\u00ebni di\u00e7ka. N\u00ebse keni ndryshuar nj\u00eb sh\u00ebrbim t\u00eb ri, ai del n\u00eb prodhim 100 her\u00eb n\u00eb or\u00eb, dhe gjith\u00e7ka \u00ebsht\u00eb mir\u00eb, at\u00ebher\u00eb njer\u00ebzit n\u00eb kompanin\u00eb tuaj mund t\u00eb thon\u00eb:<\/p>\n

\u2014 Ky \u00ebsht\u00eb nj\u00eb sh\u00ebrbim i ri! Ajo ishte gjith\u00e7ka e thjesht\u00eb, provoni t\u00eb b\u00ebni di\u00e7ka me plehrat tona.<\/em><\/p>\n

Sh\u00ebrbimi Legacy ka kuptim t\u00eb merret n\u00eb transformim kur e b\u00ebni at\u00eb me dik\u00eb, p\u00ebr shembull, n\u00ebse keni ftuar nj\u00eb konsultant t\u00eb jasht\u00ebm. T\u00eb jemi t\u00eb sinqert\u00eb, transformimi do t\u00eb trondit\u00eb gjith\u00e7ka q\u00eb mundet.<\/strong>. Ju eksperimentoni dhe nuk dini se ku do t\u00eb shkoni, cilat teknologji do t\u00eb p\u00ebrdorni dhe p\u00ebr \u00e7far\u00eb, ku dhe cilat pengesa do t\u00eb shfaqen n\u00eb procese. Prandaj, \u00ebsht\u00eb m\u00eb e leht\u00eb t\u00eb nd\u00ebrrosh di\u00e7ka t\u00eb re.<\/p>\n

N\u00ebse po e b\u00ebni gjith\u00e7ka vet\u00eb, dhe n\u00eb kompanin\u00eb tuaj nuk ka kompetenc\u00eb t\u00eb r\u00ebnd\u00ebsishme \u2014 marrim sh\u00ebrbimin e ri. N\u00ebse njihni nj\u00eb konsultant t\u00eb jasht\u00ebm dhe keni mjete \u2014 zgjidhni at\u00eb t\u00eb vjet\u00ebr.<\/p><\/blockquote>\n

\nKa sh\u00ebrbime q\u00eb p\u00ebrfaq\u00ebsojn\u00eb thjesht nj\u00eb nd\u00ebrfaqe p\u00ebr p\u00ebrdoruesit, p\u00ebr shembull, nj\u00eb sit t\u00eb thjesht\u00eb ose nj\u00eb aplikacion mobil. Por ka edhe gj\u00ebra serioze si faturimi. N\u00ebse di\u00e7ka shkon keq me faturimin \u2014 do t\u00eb jet\u00eb e v\u00ebshtir\u00eb t\u00eb merresh me t\u00eb. K\u00ebtu gjithashtu kemi nj\u00eb zgjedhje.<\/p>\n

Ne punojm\u00eb ose me nj\u00eb sh\u00ebrbim kritik<\/strong>, por p\u00ebr shkak t\u00eb tij vuajm\u00eb, ai krijon kufizime, ose punojm\u00eb me nd\u00ebrfaqen<\/strong>. Ky \u00ebsht\u00eb kriteri i dyt\u00eb i zgjedhjes. Similarisht, ka mund\u00ebsi t\u00eb angazhojm\u00eb nj\u00eb konsultant t\u00eb njohur \u2014 punojm\u00eb me variantin e r\u00ebnd\u00eb.<\/p>\n

Por edhe n\u00eb k\u00ebt\u00eb rast, nuk do t\u00eb rekomandoja ta b\u00ebni k\u00ebt\u00eb, sepse, derisa nuk ka nj\u00eb kuptim se me \u00e7far\u00eb po punoni dhe n\u00eb cilin drejtim do t\u00eb transformoni, t\u00eb merrni nj\u00eb gj\u00eb kritike dhe ta rimbusni \u2014 nuk \u00ebsht\u00eb nj\u00eb ide e mir\u00eb. Prandaj, n\u00eb k\u00ebt\u00eb rast ne preferojm\u00eb t\u00eb punojm\u00eb me nd\u00ebrfaqen, e cila nuk \u00ebsht\u00eb kritike n\u00eb rast d\u00ebshtimi.<\/p>\n

M\u00eb pas do t\u00eb shqyrtojm\u00eb ekipin e sh\u00ebrbimit<\/b>. Me ata q\u00eb merren me k\u00ebt\u00eb sh\u00ebrbim, do t\u00eb duhet t\u00eb punojm\u00eb vazhdimisht dhe t\u00eb bashk\u00ebpunojm\u00eb n\u00eb nj\u00eb kontakt t\u00eb ngusht\u00eb.<\/p>\n

Njer\u00ebzit n\u00eb ekip ndahen n\u00eb dy kategori: konservator\u00eb<\/strong> \u2014 jetojn\u00eb n\u00eb bot\u00ebn e vjet\u00ebr, ose thjesht nuk din\u00eb gj\u00eb rreth DevOps, dhe inovator\u00eb<\/strong>, q\u00eb sjellin praktikat m\u00eb moderne. T\u00eb dyt\u00ebt nuk e kuptojn\u00eb gjithmon\u00eb tem\u00ebn, por t\u00eb pakt\u00ebn jan\u00eb t\u00eb gatsh\u00ebm p\u00ebr t\u00eb.<\/p>\n

Nga nj\u00ebra an\u00eb, konservator\u00ebt jan\u00eb njer\u00ebz me p\u00ebrvoj\u00eb: jan\u00eb prej koh\u00ebsh n\u00eb kompani, kuptojn\u00eb nga fillimi deri n\u00eb fund, por nuk e din\u00eb sakt\u00ebsisht p\u00ebr praktikat. Nga ana tjet\u00ebr jan\u00eb inovator\u00ebt, t\u00eb cil\u00ebt kan\u00eb d\u00ebgjuar di\u00e7ka, por n\u00eb kompani, me siguri, nuk punojn\u00eb p\u00ebr nj\u00eb koh\u00eb t\u00eb gjat\u00eb. Me cil\u00ebt \u00ebsht\u00eb m\u00eb mir\u00eb t\u00eb punoni?<\/p>\n

Me duhet t\u00eb bashk\u00ebpunojm\u00eb me konservator\u00ebt n\u00eb \u00e7do rast, sepse ky \u00ebsht\u00eb sh\u00ebrbimi i tyre. Do na duhet t\u00eb flasim me ta, t\u00eb sqarojm\u00eb specifik\u00ebn e sh\u00ebrbimit, \u00e7far\u00eb mund t\u00eb b\u00ebjm\u00eb k\u00ebshtu dhe \u00e7far\u00eb n\u00eb t\u00eb tjet\u00ebr. Ne variojm\u00eb nga konsultimet e tyre. Sigurisht, do t\u00eb na duhet t'u japim disa detyra, sepse ata e njohin sh\u00ebrbimin e tyre m\u00eb mir\u00eb. Prandaj, \u00ebsht\u00eb e r\u00ebnd\u00ebsishme me cil\u00ebn ekip do t\u00eb kemi kontakt n\u00eb fund.<\/p>\n

\u00cbsht\u00eb logjike t\u00eb zgjedhim inovator\u00ebt n\u00eb ekip, sepse konservator\u00ebt mund t\u00eb sabotojn\u00eb pun\u00ebn.\n<\/p><\/blockquote>\n

\nN\u00eb praktik\u00eb shpesh ndodh q\u00eb njer\u00ebzit konservator\u00eb kan\u00eb p\u00ebrvoj\u00eb t\u00eb konsiderueshme, por nuk e kuptojn\u00eb si t\u00eb jetojn\u00eb m\u00eb tej. Ata thjesht frik\u00ebsohen se pas transformimit dhe riparimit t\u00eb sh\u00ebrbimit, do t\u00eb shkarkohen p\u00ebr shkak t\u00eb mos nevojshm\u00ebris\u00eb. Ndonj\u00ebher\u00eb, thjesht p\u00ebr shkak t\u00eb moskuptimit t\u00eb asaj q\u00eb ndodh, ata sabotojn\u00eb pun\u00ebn.<\/p>\n

Kam pasur nj\u00eb rast kur nj\u00eb djal\u00eb nga ekipi riparonte \u00e7do gj\u00eb, sepse p\u00ebr t\u00eb ishte m\u00eb kritike se sa ajo q\u00eb po b\u00ebjm\u00eb tani. Ne vendosim detyr\u00ebn: t\u00eb implementojm\u00eb k\u00ebt\u00eb pjes\u00eb sot \u2014 jo, n\u00eb an\u00ebn tjet\u00ebr t\u00eb bot\u00ebs ka nj\u00eb zjarr, po shkojm\u00eb ta riparojm\u00eb at\u00eb. \u00cbsht\u00eb e v\u00ebshtir\u00eb t\u00eb punosh me k\u00ebta njer\u00ebz.<\/p>\n

Njer\u00ebzit nga ekipi i konservator\u00ebve shpesh e injorojn\u00eb detyrat, ose i vonojn\u00eb deri n\u00eb momentin e fundit. Dhe n\u00ebse, zotit John Willis, b\u00ebni nj\u00eb gabim dhe u vendosni KPI p\u00ebr numrin e detyrave t\u00eb p\u00ebrfunduar, nd\u00ebrsa nj\u00eb pjes\u00eb p\u00ebr ndonj\u00eb arsye nuk \u00ebsht\u00eb e p\u00ebrfshir\u00eb n\u00eb KPI, ata nuk do t\u00eb b\u00ebjn\u00eb asgj\u00eb. N\u00eb thelb, do t\u00eb keni t\u00eb drejt\u00eb, sepse at\u00ebher\u00eb ata humbin bonusin.<\/p>\n

Me inovator\u00ebt \u00ebsht\u00eb m\u00eb e leht\u00eb \u2014 ata jan\u00eb m\u00eb lojal\u00eb.<\/strong>. Ata kan\u00eb d\u00ebgjuar tashm\u00eb di\u00e7ka, duan t\u00eb shkojn\u00eb diku, prandaj do t\u00eb ndihmojn\u00eb. Na duhen njer\u00ebz q\u00eb jan\u00eb t\u00eb gatsh\u00ebm t\u00eb vuajn\u00eb p\u00ebr nj\u00eb koh\u00eb t\u00eb par\u00eb: n\u00ebse sh\u00ebrbimi ndryshon, t\u00eb gjitha pengesat dhe v\u00ebshtir\u00ebsit\u00eb do t'i p\u00ebrjetojn\u00eb inovator\u00ebt si pionier\u00eb. Inovator\u00ebt duan gjith\u00e7ka m\u00eb t\u00eb re dhe moderne, dhe t\u00eb vuajn\u00eb.<\/p>\n

Konservator\u00ebt mund t\u00eb kthehen n\u00eb besimin tuaj m\u00eb von\u00eb. Kur t\u00eb tregoni se keni ndryshuar nj\u00eb pjes\u00eb dhe gjith\u00e7ka funksionon mir\u00eb, ka shum\u00eb mund\u00ebsi q\u00eb ata gjithashtu do t\u00eb duan ta provojn\u00eb dhe do t\u00eb pranojn\u00eb besimin e ri DevOps.<\/p>\n

\"Si
\n<\/p>\n

T\u00eb p\u00ebrmbledhim. N\u00ebse e b\u00ebjm\u00eb t\u00eb gjith\u00eb transformimin n\u00eb kompanin\u00eb ton\u00eb vet\u00eb, at\u00ebher\u00eb zgjedhim: nj\u00eb sh\u00ebrbim t\u00eb ri, preferably nj\u00eb nd\u00ebrfaqe t\u00eb thjesht\u00eb, n\u00eb m\u00ebnyr\u00eb q\u00eb t\u00eb mos vuajm\u00eb shum\u00eb nga prishet, dhe nj\u00eb ekip inovator\u00ebsh.<\/p><\/blockquote>\n

\nN\u00ebse ka mund\u00ebsi t\u00eb th\u00ebrras\u00ebsh nj\u00eb konsultant t\u00eb jasht\u00ebm, n\u00eb vend t\u00eb nj\u00eb t\u00eb riu \u2014 p\u00ebrdorim sh\u00ebrbimin e vjet\u00ebr, p\u00ebr t\u00eb cilin tashm\u00eb vuajm\u00eb. Njer\u00ebzit q\u00eb jan\u00eb angazhuar n\u00eb transformim mjaft gjat\u00eb n\u00eb vende t\u00eb ndryshme, kan\u00eb par\u00eb raste t\u00eb ndryshme dhe tashm\u00eb e din\u00eb si t\u00eb b\u00ebjn\u00eb gj\u00ebrat si\u00e7 duhet, dhe n\u00eb cilin drejtim t\u00eb shkohet.<\/p>\n

Kush \u00ebsht\u00eb i p\u00ebrfshir\u00eb?<\/h2>\n

\nNa nevojitet t\u00eb gjejm\u00eb t\u00eb gjith\u00eb ata q\u00eb kan\u00eb ndonj\u00eb lidhje me sh\u00ebrbimin: zhvilluesit, testuesit, administrator\u00ebt, sigurisht, menaxher\u00ebt dhe, ndoshta, Product Owners. Megjith\u00ebse Product Owners nuk jan\u00eb teknik\u00eb, ata kan\u00eb lidhje me sh\u00ebrbimin: marrin vendime, vendosin detyra.<\/p>\n

\"Si
\n<\/p>\n

T\u00eb gjith\u00eb ata q\u00eb marrin ndonj\u00eb vendim dhe ndikojn\u00eb n\u00eb at\u00eb q\u00eb ndodh me sh\u00ebrbimin, duhet t\u00eb gjenden, t\u00eb njihen dhe t\u00eb bisedohet me ta.<\/p><\/blockquote>\n

\nP\u00ebr \u00e7far\u00eb na nevojiten ata? Q\u00eb t\u00eb dim\u00eb me k\u00eb t\u00eb bisedojm\u00eb.<\/strong>Gjat\u00eb transformimit, kur ndryshon parimi tradicional i pun\u00ebs me sh\u00ebrbimin, do t\u00eb ket\u00eb l\u00ebkundje gjithsesi. P\u00ebr nj\u00eb koh\u00eb do t\u00eb ket\u00eb prishje, nd\u00ebrsa ne testojm\u00eb qasje t\u00eb reja. Njer\u00ebzit duhet t\u00eb jen\u00eb t\u00eb gatsh\u00ebm p\u00ebr k\u00ebt\u00eb dhe t\u00eb pranojn\u00eb.<\/p>\n

M\u00eb pas do t\u00eb duhet t\u00eb nd\u00ebrtojm\u00eb Hart\u00ebn e Vler\u00ebs dhe pa k\u00ebta njer\u00ebz nuk e nd\u00ebrton dot, sepse vet\u00ebm ata bashk\u00eb e din\u00eb plot\u00ebsisht \u00e7far\u00eb po ndodh. Nj\u00eb njeri nuk e di kurr\u00eb gjith\u00e7ka q\u00eb po ndodh me sh\u00ebrbimin.<\/p>\n

Ata do t\u00eb rekomandojn\u00eb njer\u00ebz p\u00ebr ekipin. M\u00eb von\u00eb do t\u00eb diskutojm\u00eb pse nevojitet nj\u00eb ekip i ve\u00e7ant\u00eb. N\u00eb t\u00eb do t\u00eb duhet t\u00eb marrim njer\u00ebz nga departamentet ekzistuese. Ata q\u00eb kan\u00eb lidhje me sh\u00ebrbimin, do t\u00eb mund t\u00eb rekomandojn\u00eb koleg\u00eb q\u00eb mendojn\u00eb n\u00eb drejtimin ton\u00eb, q\u00eb mund t\u00eb na ndihmojn\u00eb dhe kan\u00eb kompetenc\u00ebn q\u00eb na nevojitet.<\/p>\n

M\u00eb pas ne mbledhim t\u00eb gjith\u00eb k\u00ebta njer\u00ebz nga departamentet e ndryshme n\u00eb nj\u00eb dhom\u00eb dhe fillojm\u00eb t\u00eb nd\u00ebrtojm\u00eb Hart\u00ebn e Vler\u00ebs.<\/p>\n

Nd\u00ebrtojm\u00eb Hart\u00ebn e Vler\u00ebs<\/h2>\n

\nHarta e Vler\u00ebs \u00ebsht\u00eb nj\u00eb skem\u00eb ose hart\u00eb, e cila tregon rrjedh\u00ebn e vlerave deri te klienti.<\/strong>Ky \u00ebsht\u00eb i gjith\u00eb procesi nga ideimi i ideve deri te implementimi i tyre, duke p\u00ebrfshir\u00eb t\u00eb gjitha fazat nd\u00ebrmjet\u00ebs dhe m\u00ebnyr\u00ebn se si vlera p\u00ebrfundimisht arrin te klient\u00ebt tan\u00eb.<\/p>\n

Harta e Vler\u00ebs \u00ebsht\u00eb e nevojshme q\u00eb t\u00eb vizualizohet t\u00eb gjith\u00eb fazat e zhvillimit,<\/strong>t\u00eb lokalizohen problemet p\u00ebrmes matjeve q\u00eb jan\u00eb n\u00eb procesin aktual dhe t\u00eb fillojm\u00eb t'i eliminojm\u00eb k\u00ebto probleme, dhe t\u00eb vendosim objektivin fillestar.<\/strong>Ky \u00ebsht\u00eb vendi ku do t\u00eb fillojm\u00eb t\u00eb b\u00ebjm\u00eb di\u00e7ka me t\u00eb v\u00ebrtet\u00eb.<\/p>\n

Metrikat<\/h3>\n

\nN\u00eb literatur\u00ebn mbi Hart\u00ebn e Vler\u00ebs jan\u00eb p\u00ebrshkruar shum\u00eb metrik\u00eb t\u00eb ndryshme, por p\u00ebr fillim na mjaftojn\u00eb vet\u00ebm tre.<\/p>\n

Lead Time \u2014 vones\u00eb\/pritje<\/strong> \u2014 koha kur ne presim di\u00e7ka. P\u00ebr shembull, nj\u00eb testues pret derisa t\u00eb lirohet ndjekja p\u00ebr testet, dhe gjat\u00eb k\u00ebsaj kohe nuk mund t\u00eb b\u00ebj\u00eb asgj\u00eb.<\/p>\n

Koha e Shtuar e Vler\u00ebs \u2014 koha e pun\u00ebs s\u00eb dobishme<\/strong> \u2014 ajo q\u00eb ne kemi shpenzuar n\u00eb nj\u00eb faz\u00eb t\u00eb till\u00eb p\u00ebr t\u00eb krijuar vler\u00ebn p\u00ebrfundimtare p\u00ebr p\u00ebrdoruesin. P\u00ebr shembull, testuesi e nisi testin e tij dhe filloi t\u00eb kontrolloj\u00eb di\u00e7ka. Kjo \u00ebsht\u00eb koha e pun\u00ebs s\u00eb dobishme, kur ne me t\u00eb v\u00ebrtet\u00eb po b\u00ebjm\u00eb di\u00e7ka p\u00ebr produktin. Kjo \u00ebsht\u00eb ajo p\u00ebr t\u00eb cil\u00ebn klient\u00ebt paguajn\u00eb \u2014 p\u00ebr softin e cil\u00ebsis\u00eb.<\/p>\n

%C\/A \u2014 p\u00ebrqindja e pun\u00ebs s\u00eb pranuar. <\/strong>Ne kemi nj\u00eb faz\u00eb \u2014 zhvillimi, faza e dyt\u00eb \u2014 testimi. Sa shum\u00eb karakteristika kan\u00eb pranuar testuesit nga zhvilluesit, dhe ka k\u00ebt\u00eb p\u00ebrqindje.<\/p>\n

Dikur dukej k\u00ebshtu harta jon\u00eb.<\/p>\n

\"Si
\n
\nAjo mund t\u00eb duket ndryshe n\u00eb var\u00ebsi t\u00eb struktur\u00ebs s\u00eb organizat\u00ebs, numrit t\u00eb departamenteve dhe asaj q\u00eb b\u00ebni. Por n\u00eb p\u00ebrgjith\u00ebsi, n\u00eb hart\u00eb do t\u00eb jen\u00eb dy faza: ideja <\/strong>dhe analitika<\/strong>. N\u00eb k\u00ebt\u00eb faz\u00eb priten t\u00eb dh\u00ebna, p\u00ebr shembull, Koha e Lead 2 jav\u00eb dhe Koha e Shtuar e Vler\u00ebs 2 dit\u00eb.<\/p>\n

Metricat p\u00ebrfshijn\u00eb t\u00eb gjitha fazat.<\/p><\/blockquote>\n

\nBacklog<\/strong> \u2014 sa shum\u00eb detyra q\u00eb mbet\u00ebn pas kur analist\u00ebt i menduan ato.<\/p>\n

Zhvillimi<\/strong> \u2014 sa shum\u00eb jav\u00eb zhvilluesit prit\u00ebn sqarime p\u00ebr detyrat, ndjekjet apo pajisjet \u2014 nuk ka r\u00ebnd\u00ebsi, por ata po presin di\u00e7ka. P\u00ebr shembull, n\u00eb 4 dit\u00eb ata zbatuan nj\u00eb karakteristik\u00eb. K\u00ebtu shfaqet metrika %C\/A. Zhvilluesit mor\u00ebn nga Backlog vet\u00ebm 80% t\u00eb detyrave. Ata mendojn\u00eb se 20% e mbetura nuk kan\u00eb nj\u00eb specifikim t\u00eb mjaftuesh\u00ebm dhe i d\u00ebrguan p\u00ebr ripunim.<\/p>\n

Testimi<\/strong>. N\u00eb diagram LT \u00ebsht\u00eb caktuar 4 dit\u00eb. P\u00ebr shembull, testuesit prisnin lirimin e ndjekjes testuese, VA 2 dit\u00eb ata me t\u00eb v\u00ebrtet\u00eb ndihmonin n\u00eb testim, dhe %C\/A = 40 %. \u2014 vet\u00ebm 40 % e kodit ose karakteristikave q\u00eb zhvilluesit d\u00ebrguan, testuesit i konsideruan adekuate. E gjith\u00eb e tjerra nuk iu p\u00eblqeu p\u00ebr nj\u00eb arsye apo nj\u00eb tjet\u00ebr.<\/p>\n

Nuk do t\u00eb ndalem n\u00eb m\u00ebnyr\u00ebn se si p\u00ebr t\u00eb realizuar k\u00ebto matje, n\u00eb fund t\u00eb artikullit do t\u00eb rekomandoj literatur\u00eb p\u00ebr t\u00eb cil\u00ebn mund t\u00eb m\u00ebsoni rreth tyre.<\/p>\n

E vetmja gj\u00eb q\u00eb do t\u00eb rekomandoj \u2014 mos i besoni njer\u00ebzve q\u00eb do t\u00eb hartojn\u00eb nj\u00eb Hart\u00eb t\u00eb Vler\u00ebsimit me ju. Ata paraqesin se sa koh\u00eb zgjat proceset e ndryshme, por k\u00ebto vler\u00ebsime nuk jan\u00eb gjithmon\u00eb t\u00eb sakta, prandaj \u00ebsht\u00eb m\u00eb mir\u00eb t\u00eb matni vet\u00eb.<\/p>\n

Kemi pasur nj\u00eb rast kur shkuam n\u00eb departamentin e Operacioneve dhe pyesnim se sa koh\u00eb do t\u00eb kishte p\u00ebr t\u00eb dor\u00ebzuar nj\u00eb vefeature t\u00eb re n\u00eb prodhim. Ata na than\u00eb se 10 minuta, dhe ne menduam, pse erdh\u00ebm n\u00eb k\u00ebt\u00eb kompani? Doli se 10 minuta \u00ebsht\u00eb koha q\u00eb k\u00ebrkohet nga skripti p\u00ebr t\u00eb marr\u00eb kodin dhe p\u00ebr ta d\u00ebrguar n\u00eb server. Por para k\u00ebsaj, rrelease q\u00ebndron p\u00ebr tri dit\u00eb n\u00eb server dhe thjesht pluhuroset - n\u00eb Backlog ka nj\u00eb detyr\u00eb q\u00eb duhet t\u00eb d\u00ebrgohet. K\u00ebshtu, p\u00ebrpara faz\u00ebs s\u00eb dor\u00ebzimit, ka nj\u00eb faz\u00eb pritjeje ku projekti thjesht q\u00ebndron. Po sikur ne t\u00eb mos ishim shkuar me nj\u00eb notebook, t\u00eb mos ishim kapur me syt\u00eb p\u00ebr nj\u00eb detyr\u00eb n\u00eb Jira dhe t\u00eb mos e kishim ndjekur at\u00eb hap pas hapi, do t\u00eb mendonim se gjith\u00e7ka \u00ebsht\u00eb shk\u00eblqyer dhe s'ka asnj\u00eb problem.<\/p>\n

Prandaj, do t\u00eb duhet q\u00eb matjet t'i b\u00ebni vet\u00eb, preferohet jo vet\u00ebm nj\u00eb her\u00eb, p\u00ebr t\u00eb pasur nj\u00eb p\u00ebrmbledhje t\u00eb sakt\u00eb t\u00eb realitetit. N\u00eb var\u00ebsi t\u00eb Hart\u00ebs s\u00eb Vler\u00ebs, do t\u00eb merrni vendimin se nga ku t\u00eb filloni dhe \u00e7far\u00eb t\u00eb rregulloni si prioritet.<\/p>\n

Ekipi p\u00ebrkohsh\u00ebm<\/h2>\n

\nShum\u00eb kompani q\u00eb vendosin t\u00eb integrojn\u00eb DevOps krijojn\u00eb ekipe, por jo ekip t\u00eb p\u00ebrkohsh\u00ebm, p\u00ebrkundrazi, nj\u00eb ekip q\u00eb ekziston p\u00ebr disa vite. N\u00ebse do t'i drejtoheni sh\u00ebrbimit DevOps, ku p\u00ebrshkruhen modele t\u00eb ndryshme t\u00eb nd\u00ebrtimit t\u00eb struktur\u00ebs organizative n\u00eb DevOps, do t\u00eb kuptoni se kjo \u00ebsht\u00eb nj\u00eb anti-model.<\/p>\n

Kur ekipi DevOps ekziston vazhdimisht p\u00ebr disa vite - kjo \u00ebsht\u00eb nj\u00eb gabim i madh, sepse DevOps \u00ebsht\u00eb p\u00ebr komunikimin midis departamenteve, p\u00ebr shpejt\u00ebsin\u00eb dhe efikasitetin.<\/p><\/blockquote>\n

\nN\u00ebse ekipi ekziston midis departamenteve, vet\u00ebm p\u00ebr t\u00eb b\u00ebr\u00eb di\u00e7ka tjet\u00ebr t\u00eb ve\u00e7ant\u00eb, dhe ekziston gjat\u00eb p\u00ebr nj\u00eb koh\u00eb t\u00eb gjat\u00eb, at\u00ebher\u00eb krijon nj\u00eb barrier\u00eb t\u00eb panevojshme. Tani programuesi, p\u00ebrve\u00e7se t\u00eb shkoj\u00eb menj\u00ebher\u00eb tek administratori p\u00ebr t\u00eb zgjidhur nj\u00eb \u00e7\u00ebshtje, duhet s\u00eb pari t\u00eb drejtohet te departamenti DevOps, dhe pastaj ai do t\u00eb shkoj\u00eb m\u00eb tej.<\/p>\n

Prandaj, p\u00ebr t\u00eb filluar, \u00ebsht\u00eb e nevojshme t\u00eb krijoni nj\u00eb ekip t\u00eb p\u00ebrkohsh\u00ebm.<\/strong>. Ajo do t\u00eb ekzistoj\u00eb n\u00eb m\u00ebnyr\u00eb t\u00eb p\u00ebrkohshme p\u00ebr gjysm\u00eb viti, maksimumi nj\u00eb vit, n\u00eb var\u00ebsi t\u00eb detyr\u00ebs s\u00eb vendosur, vet\u00ebm p\u00ebr t\u00eb eliminuar nj\u00eb kufizim q\u00eb kemi zgjedhur. M\u00eb pas ajo do t\u00eb vdes\u00eb. N\u00ebse ne zgjedhim pik\u00ebn tjet\u00ebr ku na dhemb shum\u00eb, dhe kuptojm\u00eb se p\u00ebr t\u00eb na nevojitet gjithashtu nj\u00eb ekip i ve\u00e7ant\u00eb, at\u00ebher\u00eb ne do ta krijojm\u00eb s\u00ebrish. Por n\u00eb 'konsolidim' ato ekipe nuk duhet t\u00eb ekzistojn\u00eb - p\u00ebrndryshe ato thjesht do t\u00eb shkat\u00ebrrojn\u00eb komunikimin dhe do t\u00eb marrin detyra t\u00eb tjera q\u00eb nuk kan\u00eb t\u00eb b\u00ebjn\u00eb me asgj\u00eb, ve\u00e7se p\u00ebr t\u00eb b\u00ebr\u00eb di\u00e7ka. K\u00ebto detyra mund t\u00eb mos ken\u00eb lidhje me DevOps dhe me transformimin. Pse nuk duhet t'ia dor\u00ebzojm\u00eb k\u00ebt\u00eb detyr\u00eb departamenteve ekzistuese?<\/p>\n

Pse nevojitet nj\u00eb ekip p\u00ebrkoh\u00ebsor<\/h3>\n

\nKonflikti me proceset aktuale<\/strong>. Transformimi i DevOps \u00ebsht\u00eb nj\u00eb ndryshim jo vet\u00ebm n\u00eb teknologjit\u00eb dhe mjetet q\u00eb p\u00ebrdorim, por edhe n\u00eb procesin e pun\u00ebs, mendimin dhe vlerat. N\u00ebse ekipi do t\u00eb punoj\u00eb ashtu si\u00e7 \u00ebsht\u00eb m\u00ebsuar, nuk do t\u00eb jet\u00eb n\u00eb gjendje t\u00eb provoj\u00eb qasje t\u00eb tjera.<\/p>\n

K\u00ebta njer\u00ebz duhet t\u00eb jetojn\u00eb sipas rregullave t\u00eb tjera: t\u00eb injorojn\u00eb t\u00eb gjitha KPI-t\u00eb e kompanis\u00eb, sepse ata p\u00ebrpiqen t\u00eb punojn\u00eb ndryshe. Ekipet p\u00ebrkoh\u00ebsore nuk do t\u00eb dor\u00ebzojn\u00eb k\u00ebrkesa p\u00ebr t\u00eb marr\u00eb server, por do t\u00eb shkojn\u00eb direkt n\u00eb departamentin q\u00eb merret me ta, me k\u00ebrkes\u00ebn p\u00ebr t'u dh\u00ebn\u00eb asaj t\u00eb par\u00ebs at\u00eb q\u00eb iu nevojitet, sepse kjo \u00ebsht\u00eb nj\u00eb detyr\u00eb prioritare dhe sepse p\u00ebrpiqen t\u00eb jetojn\u00eb ndryshe. Ekipi ka nj\u00eb konflikt t\u00eb plot\u00eb me t\u00eb gjitha proceset aktuale. Q\u00eb metodat ekzistuese t\u00eb pun\u00ebs t\u00eb mos i pengojn\u00eb tani, dhe ata t\u00eb mos pengojn\u00eb t\u00eb tjer\u00ebt, ne izolojm\u00eb k\u00ebta njer\u00ebz, duke formuar nj\u00eb ekip t\u00eb ve\u00e7ant\u00eb.<\/p>\n

Shmangia e burokracis\u00eb n\u00eb eksperimente<\/strong>. N\u00eb ekipet p\u00ebrkoh\u00ebsore nuk ka burokraci, ata nuk plot\u00ebsojn\u00eb raporte p\u00ebr or\u00ebt e pun\u00ebs, ata nuk p\u00ebrgjigjen para menaxher\u00ebve. Ky \u00ebsht\u00eb nj\u00eb bot\u00eb krejt\u00ebsisht e ndar\u00eb, ku njer\u00ebzit jetojn\u00eb dhe mendojn\u00eb ndryshe, dhe merren me gj\u00ebra krejt\u00ebsisht t\u00eb tjera. Nuk duhet t'i pengojm\u00eb ata nj\u00eb her\u00eb tjet\u00ebr.<\/p>\n

Puna e pand\u00ebrprer\u00eb mbi sh\u00ebrbimin<\/strong>. N\u00eb pik\u00ebn e par\u00eb ne zgjodh\u00ebm di\u00e7ka mbi t\u00eb cil\u00ebn do t\u00eb eksperimentojm\u00eb. Eksperimentet dhe k\u00ebrkimi i m\u00ebnyrave p\u00ebr t\u00eb punuar m\u00eb mir\u00eb \u00ebsht\u00eb i mir\u00eb, por ne duam t\u00eb prodhojm\u00eb gjithashtu ve\u00e7ori. N\u00ebse e gjith\u00eb ekipi merret me transformimin n\u00eb vend t\u00eb ve\u00e7orive, at\u00ebher\u00eb ne do t\u00eb fillojm\u00eb t\u00eb humbasim t\u00eb ardhurat, gabimet do t\u00eb q\u00ebndrojn\u00eb p\u00ebr nj\u00eb koh\u00eb t\u00eb gjat\u00eb - kjo nuk \u00ebsht\u00eb ajo q\u00eb na nevojitet. Krijimi i nj\u00eb ekipi t\u00eb p\u00ebrkohsh\u00ebm lejon t\u00eb eksperimentosh pa ndalur pun\u00ebn mbi produktin.<\/p>\n

Mos ndiheni n\u00eb pun\u00eb,<\/strong>. Kjo s\u00ebrish \u00ebsht\u00eb p\u00ebr produktin. Duhet shum\u00eb koh\u00eb q\u00eb ekipi t\u00eb provoj\u00eb mjete t\u00eb tjera e t\u00eb tjera. Q\u00eb njer\u00ebzit t\u00eb m\u00ebsojn\u00eb mjetet, t\u00eb fillojn\u00eb t'i zbatojn\u00eb dhe t'i p\u00ebrdorin si\u00e7 duhet, do t\u00eb marr\u00eb t\u00eb pakt\u00ebn gjasht\u00eb muaj. N\u00ebse ata merrejn\u00eb edhe me produktin \u2014 gjasht\u00eb muaj do t\u00eb zgjaten shum\u00eb. N\u00ebse njer\u00ebzit punojn\u00eb me produktin, ata p\u00ebrs\u00ebri po punojn\u00eb me proceset e vjetra \u2014 kjo nuk na nevojitet.<\/p>\n

Prandaj nga departamentet e ndryshme ne ve\u00e7ojm\u00eb njer\u00ebz n\u00eb nj\u00eb ekip t\u00eb ve\u00e7ant\u00eb q\u00eb do t\u00eb merret me transformimin e sh\u00ebrbimit. Si rezultat, sh\u00ebrbimi funksionon, vazhdon t\u00eb zhvillohet, dhe n\u00eb t\u00eb nj\u00ebjt\u00ebn koh\u00eb ne realizojm\u00eb disa eksperimente mbi t\u00eb.<\/p>\n

Ekipi temporal merret vet\u00ebm me transformimin DevOps \u2014 eliminimin e atij kufizimi q\u00eb kemi gjetur, dhe asgj\u00eb m\u00eb shum\u00eb.<\/p><\/blockquote>\n

\nEkipi p\u00ebrb\u00ebhet nga njer\u00ebz universale,<\/strong>. Kjo do t\u00eb thot\u00eb se ne nuk kemi marr\u00eb vet\u00ebm zhvillues. Nuk erdh\u00ebm n\u00eb sh\u00ebrbim dhe nuk e mor\u00ebm half-e ekipit \u2014 jo, ne mor\u00ebm njer\u00ebz nga departamente t\u00eb ndryshme,<\/strong>. Disa pika m\u00eb par\u00eb ne gjet\u00ebm departamente dhe punonj\u00ebs t\u00eb ndrysh\u00ebm q\u00eb kan\u00eb lidhje me sh\u00ebrbimin e transformuar. Nga ata ne nd\u00ebrtojm\u00eb ekipin, sepse ai duhet t\u00eb jet\u00eb universal \u2014 ne do t\u00eb ndryshojm\u00eb procesin e testimit, procesin e zhvillimit dhe procesin e sh\u00ebrbimit. Duhet kompetenca t\u00eb ndryshme.<\/p>\n

P\u00ebr zakonisht, ne marrim nj\u00eb zhvillues, nj\u00eb testues dhe nj\u00eb inxhinier \u2014 secilin nga nj\u00eb dhe s\u00eb bashku me ta inventarizojm\u00eb nj\u00eb zgjidhje, e cila lejon t\u00eb jetojm\u00eb ndryshe.<\/p>\n

Preferohet q\u00eb k\u00ebta njer\u00ebz t\u00eb ken\u00eb autoritet n\u00eb organizat\u00eb,<\/strong>. Mund t\u00eb nevojitet t\u00eb marr\u00ebsh nj\u00eb konservator, ndon\u00ebse nuk d\u00ebshirohet. N\u00ebse kemi nj\u00eb kompani t\u00eb madhe, nuk t\u00eb gjith\u00eb do t\u00eb besojn\u00eb n\u00eb projektin ton\u00eb, dhe ndokush mund t\u00eb fut\u00eb pengesa, p.sh., t\u00eb mos alokoj\u00eb nj\u00eb stand. Aty do t\u00eb nevojitet \"autoritete\" \u2014 nj\u00eb njeri i respektuar me p\u00ebrvoj\u00eb t\u00eb madhe, i cili ka fituar nj\u00eb marr\u00ebdh\u00ebnie t\u00eb mir\u00eb me koleg\u00ebt. Autoriteti i punonj\u00ebsit n\u00eb ekip do t\u00eb thjeshtoj\u00eb detyr\u00ebn dhe pun\u00ebn e ekipit temporal. Njer\u00ebzit do t\u00eb mendojn\u00eb:<\/p>\n

\u2014 Ah, ky \u00ebsht\u00eb ai djal\u00eb i shk\u00eblqyer, q\u00eb t\u00eb gjith\u00eb e njohim dhe e duam, q\u00eb \u00ebsht\u00eb angazhuar \u2014 duket se n\u00eb DevOps ka di\u00e7ka q\u00eb meritohet t\u00eb shikohet!<\/em><\/p>\n

Vendosim q\u00ebllimin,<\/h2>\n

\nMblodh\u00ebm njer\u00ebzit, zgjodh\u00ebm sh\u00ebrbimin, shik\u00ebm kufizimet, p\u00ebrcaktuam cil\u00ebt njer\u00ebz do t\u00eb ndikojm\u00eb. Tani duhet t\u00eb vendosim nj\u00eb q\u00ebllim dhe ai duhet t\u00eb jet\u00eb t\u00ebr\u00ebsisht sip\u00ebr SMART,<\/strong> \u2014 ashtu si\u00e7 na p\u00eblqen.<\/p>\n

Specifik \u2014 specifike<\/strong>.<\/p>\n

Mat\u00ebr \u2014 e matshme<\/strong>. Ky \u00ebsht\u00eb nj\u00eb pik\u00eb shum\u00eb e r\u00ebnd\u00ebsishme e SMART. N\u00ebse nuk mund t\u00eb matni di\u00e7ka, at\u00ebher\u00eb nuk mund ta ndryshoni dhe nuk kuptoni se \u00e7far\u00eb dhe si e keni b\u00ebr\u00eb m\u00eb mir\u00eb ose m\u00eb keq.<\/p>\n

Arritsh\u00ebm \u2014 e arritshme<\/strong>. B\u00ebni nj\u00eb p\u00ebrshtatje p\u00ebr specifik\u00ebn tuaj. N\u00ebse jeni nj\u00eb kompani enterprise me nj\u00eb histori t\u00eb gjat\u00eb dhe nj\u00eb ngarkes\u00eb t\u00eb madhe detyrash, e cila l\u00ebshon nj\u00eb version produkti \u00e7do vit, at\u00ebher\u00eb nuk do t\u00eb mund t\u00eb arrini l\u00ebshimin e versioneve t\u00eb reja t\u00eb produktit \u00e7do or\u00eb brenda gjasht\u00eb muajve. K\u00ebshtu nuk funksionon. Prandaj, vendosni nj\u00eb q\u00ebllim t\u00eb arritsh\u00ebm brenda nj\u00eb kohe t\u00eb pranueshme.<\/p>\n

Relevante \u2014 e r\u00ebnd\u00ebsishme. <\/strong>Eliminojm\u00eb vet\u00ebm at\u00eb kufizim q\u00eb n\u00eb t\u00eb v\u00ebrtet\u00eb synon q\u00ebllimet tona aktuale.<\/p>\n

E kufizuar n\u00eb koh\u00eb \u2014 e kufizuar nga koha<\/strong>. N\u00ebse nuk ka afat \u2014 ekipi do t\u00eb merret me \u00e7far\u00ebdo: provuar 15 teknologji n\u00eb vend t\u00eb 3, shkruar raporte t\u00eb m\u00ebdha, kryer k\u00ebrkime t\u00eb padobishme, p\u00ebrmir\u00ebsuar realizimin e tyre deri n\u00eb shk\u00eblqim, kur q\u00ebllimi \u00ebsht\u00eb arritur tashm\u00eb.<\/p>\n

Ne e marrim q\u00ebllimin pik\u00ebrisht me an\u00eb t\u00eb Value Stream Map \u2014 p\u00ebrs\u00ebri mblidhni gjith\u00eb njer\u00ebzit dhe vizatoni. Por tani, mbi baza t\u00eb Value Stream Map t\u00eb m\u00ebparshme, vizatojm\u00eb at\u00eb q\u00eb duam t\u00eb arrijm\u00eb.<\/p>\n

\"Si
\n
\nIdentifikojm\u00eb nj\u00eb kufizim, q\u00eb do t\u00eb eliminojm\u00eb tani \u2014 me at\u00eb do t\u00eb merret ekipi. P\u00ebr shembull, kam marr\u00eb pritjen nga l\u00ebshimi i gatsh\u00ebm deri n\u00eb implementimin n\u00eb prodhim \u2014 kjo \u00ebsht\u00eb kufizimi m\u00eb i zakonsh\u00ebm p\u00ebr t\u00eb cilin njer\u00ebzit i drejtohen konsulent\u00ebve.<\/p>\n

N\u00eb baz\u00eb t\u00eb k\u00ebsaj vendosim detyr\u00ebn: duam q\u00eb pritja mes l\u00ebshimit t\u00eb gatsh\u00ebm dhe daljes n\u00eb treg t\u00eb jet\u00eb maksimum nj\u00eb or\u00eb.<\/p>\n

Shembuj detyrash.<\/p>\n