{"id":41818,"date":"2020-02-16T20:46:08","date_gmt":"2020-02-16T17:46:08","guid":{"rendered":"https:\/\/prohoster.info\/blog\/blog_prohoster\/sem-arhetipov-prevrashheniya-po-princzipam-devops"},"modified":"2020-02-16T20:46:08","modified_gmt":"2020-02-16T17:46:08","slug":"sem-arhetipov-prevrashheniya-po-princzipam-devops","status":"publish","type":"post","link":"https:\/\/prohoster.info\/lv\/blog\/sem-arhetipov-prevrashheniya-po-princzipam-devops","title":{"rendered":"Septi\u0146i p\u0101rveides arhetipi, balstoties uz DevOps principiem","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

Jaut\u0101jums \u00abk\u0101 ieviest devops\u00bb jau gadiem ilgi ir aktu\u0101ls, ta\u010du labu materi\u0101lu ir maz. Da\u017ek\u0101rt j\u016bs k\u013c\u016bstat par rekl\u0101mas m\u0113r\u0137i ne \u012bpa\u0161i gudriem konsultantiem, kuriem j\u0101izp\u0101rdod savs laiks, neatkar\u012bgi no t\u0101, k\u0101. Da\u017ek\u0101rt tas ir neskaidrs, \u013coti visp\u0101r\u012bgs runas veids par to, k\u0101 megakorpor\u0101ciju ku\u0123i piedal\u0101s visuma pla\u0161umos. Rodas jaut\u0101jums: ko m\u0113s no t\u0101 ieg\u016bstam? God\u0101tais autors, vai varat skaidr\u0101k uzskait\u012bt savas idejas?<\/p>\n

Tas viss notiek t\u0101d\u0113\u013c, ka re\u0101las prakses un izpratnes par uz\u0146\u0113muma kult\u016bras transform\u0101ciju rezult\u0101tiem nav uzkr\u0101jies daudz. Kult\u016bras izmai\u0146as ir ilgsto\u0161as lietas, kuru rezult\u0101ti nepar\u0101d\u012bsies ne p\u0113c ned\u0113\u013cas, ne m\u0113ne\u0161a. Mums nepiecie\u0161ams k\u0101ds pietiekami pieredz\u0113jis, redz\u0113jis, k\u0101 uz\u0146\u0113mumi tika veidoti un sabrukti gadu gait\u0101.<\/p>\n

\"Septi\u0146i
\n
\nD\u017eons Viliss<\/b> \u2014 viens no DevOps t\u0113viem. D\u017eonam ir desmitiem gadu pieredzes darb\u0101 ar lielu skaitu uz\u0146\u0113mumu. Laika gait\u0101 D\u017eons ir paman\u012bjis specifiskus mode\u013cus, kas sastopami katra uz\u0146\u0113muma darb\u012bb\u0101. Izmantojot \u0161os arhetipus, D\u017eons vada uz\u0146\u0113mumus uz patieso DevOps transform\u0101cijas ce\u013cu. Vair\u0101k par \u0161iem arhetipiem \u2014 D\u017eona lekcijas tulkojum\u0101 no konferences DevOops 2018.
\n<\/a><\/noindex>
\n

\r\n \"Atska\u0146ot\r\n
<\/div>\r\n <\/div><\/center><\/p>\n

Par lektoru:<\/b><\/p>\n

Vair\u0101k nek\u0101 35 gadi IT vad\u012bb\u0101, piedal\u012bjies OpenCloud priek\u0161te\u010da izveid\u0113 Canonical, piedal\u012bjies 10 startapos, divus no kuriem p\u0101rdevis Dell un Docker. Pa\u0161laik ir DevOps un Digital Practices viceprezidents SJ Technologies.<\/p>\n

T\u0101l\u0101k seko st\u0101st\u012bjums no D\u017eona.<\/b><\/p>\n

Mani sauc D\u017eons Viliss, un mani visviegl\u0101k var atrast Twitter\u012b, @botchagalupe<\/a><\/noindex>. Tas pats segv\u0101rds ir ar\u012b man\u0101 Gmail un GitHub. Un caur \u0161o saiti<\/a><\/noindex> j\u016bs varat atrast videoierakstus par man\u0101m lekcij\u0101m un prezent\u0101cij\u0101m.<\/p>\n

Man ir daudz tik\u0161an\u0101s ar CIO da\u017e\u0101du lielu uz\u0146\u0113mumu. Vi\u0146i \u013coti bie\u017ei s\u016bdzas, ka nesaprot, kas ir DevOps, bet visi, kas m\u0113\u0123ina to izskaidrot, run\u0101 par kaut ko savu. V\u0113l viena bie\u017ea s\u016bdz\u012bba \u2014 DevOps nestr\u0101d\u0101, lai gan direktori \u0161\u0137iet, ka visu dara, k\u0101 vi\u0146iem skaidrots. Run\u0101jam par lieliem uz\u0146\u0113mumiem, kuriem ir vair\u0101k nek\u0101 simts gadi. Sarun\u0101joties ar vi\u0146iem, es secin\u0101ju, ka daudz\u0101m probl\u0113m\u0101m lab\u0101k noder nevis augst\u0101s tehnolo\u0123ijas, bet sal\u012bdzino\u0161i zemas tehnolo\u0123ijas risin\u0101jumi. Ned\u0113\u013c\u0101m ilgi es vienk\u0101r\u0161i sarun\u0101jos ar cilv\u0113kiem no da\u017e\u0101diem departementiem. Tas, ko j\u016bs redzat pirmaj\u0101 att\u0113l\u0101 \u0161aj\u0101 ierakst\u0101 \u2014 ir mans p\u0113d\u0113jais projekts, istaba t\u0101 izskat\u012bj\u0101s p\u0113c trim darba dien\u0101m.<\/p>\n

Kas ir DevOps?<\/h2>\n

\nPaties\u012bb\u0101, ja pajaut\u0101jat 10 da\u017e\u0101diem cilv\u0113kiem, vi\u0146i dos 10 da\u017e\u0101das atbildes. Bet interesanti ir tas, ka visas \u0161\u012bs desmit atbildes b\u016bs pareizas. Nepareiza atbilde te nav. Es jau diezgan dzi\u013ci esmu nodarbojies ar DevOps apm\u0113ram 10 gadus, biju pirmais amerik\u0101nis pirmaj\u0101 DevOpsDay. Ne teik\u0161u, ka esmu gudr\u0101ks par visiem, kas nodarbojas ar DevOps, bet gan maz ticams, ka k\u0101ds cits ir ieguld\u012bjis tikpat daudz p\u016b\u013cu. Es uzskatu, ka DevOps rodas tad, kad cilv\u0113ku kapit\u0101ls savienojas ar tehnolo\u0123ij\u0101m. Bie\u017ei aizmirstam par cilv\u0113ka dimensiju, lai gan run\u0101jam daudz par kult\u016br\u0101m. <\/p>\n

\"Septi\u0146i
\n
\nTagad mums ir daudz datu, piecus gadus ilgsto\u0161i akad\u0113misku p\u0113t\u012bjumu, teoriju p\u0101rbaude ir izveidota industri\u0101l\u0101 m\u0113rog\u0101. \u0160ie p\u0113t\u012bjumi mums saka sekojo\u0161o: ja organiz\u0101cijas kult\u016br\u0101 savieno da\u017eus uzved\u012bbas modelus, var ieg\u016bt 2000 reizes \u0101tr\u0101ku izaugsmi. \u0160im pa\u0101trin\u0101jumam atbilst t\u0101ds pats uzlabojums notur\u012bb\u0101. Tas ir kvantitat\u012bvs m\u0113r\u012bjums tam, k\u0101du priek\u0161roc\u012bbu DevOps var sniegt jebkuram uz\u0146\u0113mumam. Pirms diviem gadiem es st\u0101st\u012bju par DevOps Fortune 5000 uz\u0146\u0113muma \u0123ener\u0101ldirektoram. Kad es gatavoju prezent\u0101ciju, es \u013coti uztraucos, jo man bija j\u0101izkl\u0101sta sava daudzu gadu pieredze 5 min\u016bt\u0113s. <\/p>\n

Kopum\u0101 es sniedzu sekojo\u0161o defin\u012bciju DevOps<\/b>: tas ir prakse un mode\u013cu kopums, kas \u013cauj p\u0101rv\u0113rst cilv\u0113ku kapit\u0101lu augstas ra\u017e\u012bguma organiz\u0101cijas kapacit\u0101t\u0113. Piem\u0113rs \u2014 k\u0101 p\u0113d\u0113jos 50 vai 60 gados str\u0101d\u0101 Toyota.<\/p>\n

\"Septi\u0146i
\n
\n(\u0160aj\u0101 un turpm\u0101kaj\u0101s sh\u0113m\u0101s tiek sniegtas nevis k\u0101 atsauces materi\u0101ls, bet k\u0101 ilustr\u0101cija. To saturs at\u0161\u0137irsies katrai jaunai uz\u0146\u0113mumam. Tom\u0113r att\u0113lu var atsevi\u0161\u0137i apl\u016bkot un palielin\u0101t pa \u0161o saiti.)<\/a><\/noindex><\/i><\/p>\n

Viena no veiksm\u012bg\u0101kaj\u0101m \u0161\u0101d\u0101m praks\u0113m ir value stream mapping<\/b>. Par to ir uzrakst\u012btas vair\u0101kas labas gr\u0101matas, no kur\u0101m veiksm\u012bg\u0101kais autors ir Karen Martin. Bet p\u0113d\u0113j\u0101 gada laik\u0101 esmu non\u0101cis pie secin\u0101juma, ka pat \u0161\u012b pieeja ir p\u0101r\u0101k augsto tehnolo\u0123iju. Tam noteikti ir daudz priek\u0161roc\u012bbu, kuras esmu daudz izmantojis. Bet, kad izpilddirektors jaut\u0101, k\u0101p\u0113c vi\u0146a uz\u0146\u0113mums nevar p\u0101riet uz jaun\u0101m slied\u0113m, par value stream mapping run\u0101t v\u0113l ir p\u0101ragri. Ir daudz b\u016btisk\u0101ku jaut\u0101jumu, uz kuriem iepriek\u0161 nepiecie\u0161ams atrast atbildes. <\/p>\n

Man \u0161\u0137iet, ka daudzu manu kol\u0113\u0123u k\u013c\u016bda ir t\u0101, ka vi\u0146i vienk\u0101r\u0161i dod uz\u0146\u0113mumam piecu punktu vadl\u012bnijas, un p\u0113c tam p\u0113c se\u0161iem m\u0113ne\u0161iem atgrie\u017eas un skat\u0101s, kas noticis. Pat labai sh\u0113mai k\u0101 value stream mapping ir, t\u0101 teikt, miru\u0161\u0101s zonas (blind spots). P\u0113c simtiem interviju ar da\u017e\u0101du uz\u0146\u0113mumu direktoriem esmu izstr\u0101d\u0101jis noteiktu modeli, kas \u013cauj sadal\u012bt probl\u0113mu sast\u0101vda\u013c\u0101s, un tagad m\u0113s pa k\u0101rtai apspried\u012bsim katru no \u0161\u012bm sast\u0101vda\u013c\u0101m. Pirms jebk\u0101du tehnolo\u0123isku risin\u0101jumu izmanto\u0161anas, es izmantoju \u0161o modeli, un rezult\u0101t\u0101 man vis\u0101s sien\u0101s izn\u0101k t\u0101das sh\u0113mas. Nesen str\u0101d\u0101ju ar vienu ieguld\u012bjumu fondu, un man rezult\u0101t\u0101 bija 100-150 \u0161\u0101du sh\u0113mu.<\/p>\n

Slikta kult\u016bra katru labu pieeju ap\u0113d brokast\u012bs<\/h2>\n

\nGalven\u0101 doma ir t\u0101da: nek\u0101das Lean, Agile, SAFE un DevOps nepal\u012bdz\u0113s, ja pati organiz\u0101cijas kult\u016bra ir slikta. Tas ir t\u0101pat k\u0101 lekt dzi\u013cum\u0101 bez skafandra vai oper\u0113t bez rentgena att\u0113la. Citiem v\u0101rdiem sakot, p\u0101rfr\u0101z\u0113jot Drukeru un Demingu: slikta organiz\u0101cijas kult\u016bra noraks jebkuru labu sist\u0113mu un neap\u0113d\u012bs. <\/p>\n

Lai atrisin\u0101tu \u0161o galveno probl\u0113mu, ir j\u0101veic \u0161\u0101di so\u013ci:<\/p>\n

    \n
  1. Make All Work Visible:<\/b> j\u0101padara viss darbs redzams. Ne taj\u0101 noz\u012bm\u0113, ka tam oblig\u0101ti j\u0101par\u0101d\u0101s uz k\u0101da ekr\u0101na, bet taj\u0101 noz\u012bm\u0113, ka tam j\u0101b\u016bt nov\u0113rojamam.<\/li>\n
  2. Consolidate Work Management Systems:<\/b> ir nepiecie\u0161ams konsolid\u0113t vad\u012bbas sist\u0113mas. Attiec\u012bb\u0101 uz \u201ecil\u0161u\u201d zin\u0101\u0161an\u0101m un institucion\u0101laj\u0101m zin\u0101\u0161an\u0101m 9 no 10 gad\u012bjumiem \u0161aurs posms ir cilv\u0113ki. Gr\u0101mat\u0101 \u201ePhoenix Project\u201d<\/a><\/noindex> probl\u0113ma bija vien\u0101 vien\u012bg\u0101 cilv\u0113k\u0101, Brent\u0101, d\u0113\u013c kura projekts kav\u0113j\u0101s tr\u012bs gadus. Un uz t\u0101diem \u201eBrentiem\u201d es sastopos visur. Lai atrisin\u0101tu \u0161os \u0161auros posmus, es izmantoju \u0161os divus punktus m\u016bsu sarakst\u0101. <\/li>\n
  3. Theory of Constraints Methodology:<\/b> ierobe\u017eojumu teorija.<\/li>\n
  4. Collaboration hacks:<\/b> sadarb\u012bbas haiki. <\/li>\n
  5. Toyota Kata (Coaching Kata<\/a><\/noindex>):<\/b> par Toyota Kata es daudz nerun\u0101\u0161u. Ja interes\u0113, man\u0101 GitHub\u0101 ir prezent\u0101cijas<\/a><\/noindex> gandr\u012bz par katru no \u0161\u012bm t\u0113m\u0101m. <\/li>\n
  6. Market Oriented Organization:<\/b> tirgum orient\u0113ta organiz\u0101cija.<\/li>\n
  7. Shift-left auditors:<\/b> audits agr\u012bn\u0101s cikla stadij\u0101s.<\/li>\n<\/ol>\n

    \"Septi\u0146i
    \n
    \nEs s\u0101ku darbu ar organiz\u0101ciju \u013coti vienk\u0101r\u0161i: ieeju uz\u0146\u0113mum\u0101 un run\u0101ju ar darbiniekiem. K\u0101 redzam, nav nepiecie\u0161amas augstas tehnolo\u0123ijas. Viss, kas nepiecie\u0161ams, ir kaut kas, uz k\u0101 rakst\u012bt. Es sav\u0101cu vair\u0101kas komandas vien\u0101 telp\u0101 un anal\u012bz\u0113ju to, ko vi\u0146i man saka, no maniem 7 arhetipiem. P\u0113c tam es vi\u0146iem dodu mar\u0137ieri un l\u016bdzu uz d\u0113\u013ca uzrakst\u012bt visu to, ko vi\u0146i l\u012bdz \u0161im teica bals\u012b. Parasti \u0161\u0101da veida san\u0101ksm\u0113s ir viens cilv\u0113ks, kur\u0161 visu pieraksta, un lab\u0101kaj\u0101 gad\u012bjum\u0101 vi\u0146am izdodas pierakst\u012bt 10% diskusijas. Ar manu metodi \u0161o r\u0101d\u012bt\u0101ju izdodas paaugstin\u0101t l\u012bdz kaut kur 40%. <\/p>\n

    \"Septi\u0146i
    \n
    \n(Atsevi\u0161\u0137i \u0161o ilustr\u0101ciju var     apskat\u012bt pa saiti<\/a><\/noindex>)<\/i><\/p>\n

    Mans pieg\u0101jiens balst\u0101s uz Viljama \u0160neidera (William Schneider, The Reengineering Alternative<\/a><\/noindex>). Pieejas pamat\u0101 ir doma, ka jebkuru organiz\u0101ciju var sadal\u012bt \u010detros kvadrantos. \u0160\u012b sh\u0113ma man parasti ir rezult\u0101ts darb\u0101 ar t\u0101m simtiem citu sh\u0113mu, kas rodas, analiz\u0113jot organiz\u0101ciju. Pie\u0146emsim, ka mums ir organiz\u0101cija ar augstu kontroli, bet zemu kompetenci. Tas ir \u0101rk\u0101rt\u012bgi nev\u0113lams variants: kad visi staig\u0101 pa st\u012bg\u0101m, bet neviens nezina, ko vajag dar\u012bt. <\/p>\n

    K\u0101 lab\u0101ka alternat\u012bva ar augstu kontroli un kompetenci. Ja uz\u0146\u0113mums g\u016bst pe\u013c\u0146u, iesp\u0113jams, ka DevOps vi\u0146am nav vajadz\u012bgs. Visinteresant\u0101k ir str\u0101d\u0101t ar uz\u0146\u0113mumu, kuram ir augsta kontroles pak\u0101pe, zema kompetence un sadarb\u012bba, ta\u010du taj\u0101 pa\u0161\u0101 laik\u0101 augsta kult\u016bras l\u012bme\u0146a (cultivation) pras\u012bbas. Tas noz\u012bm\u0113, ka uz\u0146\u0113mum\u0101 ir daudz cilv\u0113ku, kuriem pat\u012bk tur str\u0101d\u0101t, un darbinieku rot\u0101cija ir zema. <\/p>\n

    \"Septi\u0146i
    \n
    \n(Atsevi\u0161\u0137i \u0161o ilustr\u0101ciju var     apskat\u012bt pa saiti<\/a><\/noindex>)<\/i><\/p>\n

    Man \u0161\u0137iet, ka metodes ar stingri noteikt\u0101m rekomend\u0101cij\u0101m galu gal\u0101 trauc\u0113 sasniegt paties\u012bbu. Jo \u012bpa\u0161i value stream mapping ir daudz noteikumu par to, k\u0101 j\u0101struktur\u0113 inform\u0101cija. Agr\u012bnos darba posmos, par kuriem es run\u0101ju, \u0161ie noteikumi nevienam nav nepiecie\u0161ami. Ja cilv\u0113ks ar mar\u0137ieri rok\u0101s apraksta uz galda eso\u0161o situ\u0101ciju uz\u0146\u0113mum\u0101 \u2014 tas ir lab\u0101kais veids, k\u0101 izprast situ\u0101ciju. \u0160\u0101da inform\u0101cija nenon\u0101k l\u012bdz direktoriem. \u0160aj\u0101 br\u012bd\u012b b\u016btu mu\u013c\u0137\u012bgi p\u0101rtraukt cilv\u0113ku un teikt, ka vi\u0146\u0161 nepareizi uzvilcis k\u0101du bultu. \u0160aj\u0101 posm\u0101 lab\u0101k izmantot vienk\u0101r\u0161as pamatnoteikumu, piem\u0113ram: daudzniveles abstrakciju var izveidot, vienk\u0101r\u0161i izmantojot kr\u0101sainus mar\u0137ierus. <\/p>\n

    Atk\u0101rtoju, nek\u0101du augsto tehnolo\u0123iju. Ar melno mar\u0137ieri att\u0113lo objekt\u012bvo realit\u0101ti, k\u0101 viss darbojas. Ar sarkano mar\u0137ieri cilv\u0113ki atz\u012bm\u0113, kas tie\u0161i vi\u0146iem eso\u0161aj\u0101 situ\u0101cij\u0101 nepat\u012bk. Ir svar\u012bgi, ka \u0161o inform\u0101ciju raksta vi\u0146i, nevis es. Kad p\u0113c sapulces es dodos pie IT direktora, es neiesniedzu sarakstu ar 10 liet\u0101m, kas j\u0101pabeidz. Es cen\u0161os atrast saiknes starp to, ko saka cilv\u0113ki no uz\u0146\u0113muma, un eso\u0161aj\u0101m p\u0101rbaud\u012btaj\u0101m sh\u0113m\u0101m. Beig\u0101s ar zilo mar\u0137ieri tiek pied\u0101v\u0101tas iesp\u0113jam\u0101s risin\u0101jumu probl\u0113mas. <\/p>\n

    \"Septi\u0146i
    \n
    \n(Atsevi\u0161\u0137i \u0161o ilustr\u0101ciju var     apskat\u012bt pa saiti<\/a><\/noindex>)<\/i><\/p>\n

    \u0160\u0101da pieeja \u0161obr\u012bd ir att\u0113lota augst\u0101k. \u0160\u012b gada s\u0101kum\u0101 es str\u0101d\u0101ju ar vienu banku. Dro\u0161\u012bbas noda\u013cas darbinieki bija p\u0101rliecin\u0101ti, ka vi\u0146iem nedr\u012bkst ierasties uz pras\u012bbu un projekt\u0113\u0161anas (design and requirement reviews) p\u0101rbaud\u0113m. <\/p>\n

    \"Septi\u0146i
    \n
    \n(Atsevi\u0161\u0137i \u0161o ilustr\u0101ciju var     apskat\u012bt pa saiti<\/a><\/noindex>)<\/i><\/p>\n

    P\u0113c tam m\u0113s run\u0101j\u0101m ar cilv\u0113kiem no cit\u0101m noda\u013c\u0101m un noskaidroj\u0101m, ka kaut kur pirms 8 gadiem programmat\u016bras izstr\u0101d\u0101t\u0101ji atlaida dro\u0161\u012bbas darbiniekus, jo tie pal\u0113nin\u0101ja darbu. Un tad tas p\u0101rv\u0113rt\u0101s par liegumu, kas tika uztverams k\u0101 pa\u0161saprotams. Lai gan paties\u012bb\u0101 nek\u0101du liegumu nebija. <\/p>\n

    M\u016bsu tik\u0161an\u0101s norit\u0113ja \u013coti sare\u017e\u0123\u012bt\u0101 veid\u0101: tr\u012bs stundu laik\u0101 piecas da\u017e\u0101das komandas nesp\u0113ja man izskaidrot, kas notiek starp kodu un b\u016bvniec\u012bbu. Un tas \u0161\u0137iet visvienk\u0101r\u0161\u0101k\u0101 lieta. Liel\u0101k\u0101 da\u013ca DevOps konsultantu uzreiz pie\u0146em, ka tas jau visiem ir zin\u0101ms. <\/p>\n

    P\u0113c tam cilv\u0113ks, kur\u0161 atbild\u0113ja par IT regul\u0113jumu (IT governance), kas \u010detrus stundas klus\u0113ja, p\u0113k\u0161\u0146i ats\u0101ka run\u0101t, kad non\u0101c\u0101m pie vi\u0146a t\u0113mas, un aiz\u0146\u0113ma m\u016bs v\u0113l uz sam\u0113r\u0101 ilgu laiku. Beig\u0101s es vi\u0146am jaut\u0101ju, ko vi\u0146\u0161 dom\u0101 par tik\u0161anos, un es nekad neaizmirst\u012b\u0161u vi\u0146a atbildi. Vi\u0146\u0161 teica: \u00abAgr\u0101k es dom\u0101ju, ka m\u016bsu bank\u0101 ir tikai divi programmat\u016bras pieg\u0101des veidi, bet tagad es zinu, ka tos ir pieci, un par trim es pat nenojautu.\u00bb <\/p>\n

    \"Septi\u0146i
    \n
    \n(Atsevi\u0161\u0137i \u0161o ilustr\u0101ciju var     apskat\u012bt pa saiti<\/a><\/noindex>)<\/i><\/p>\n

    P\u0113d\u0113j\u0101 tik\u0161an\u0101s \u0161aj\u0101 bank\u0101 bija ar komandu, kas nodarbojas ar invest\u012bciju programmat\u016bru. Tie\u0161i ar \u0161o komandu izr\u0101d\u012bj\u0101s, ka rakst\u012bt sh\u0113mas ar mar\u0137ieri uz pap\u012bra ir lab\u0101k nek\u0101 uz t\u0101feles un pat lab\u0101k nek\u0101 uz viedt\u0101feles. <\/p>\n

    \"Septi\u0146i
    \n
    \nFotogr\u0101fijas, kuras j\u016bs redzat, ir tas, k\u0101 izskat\u012bj\u0101s viesn\u012bcas konferen\u010du z\u0101le m\u016bsu tik\u0161an\u0101s ceturtaj\u0101 dien\u0101. Un \u0161\u012bs sh\u0113mas m\u0113s izmantoj\u0101m, lai mekl\u0113tu paraugus, t.i., arhetipus. <\/p>\n

    T\u0101tad, es uzdodu jaut\u0101jumus darbiniekiem, vi\u0146i ieraksta atbildes ar tr\u012bs kr\u0101su mar\u0137ieriem (melnu, sarkanu un zilu). Vi\u0146u atbildes es analiz\u0113ju p\u0113c arhetipiem. Tagad apskat\u012bsim visus arhetipus sec\u012bb\u0101. <\/p>\n

    1. Make All Work Visible: Padar\u012bt visu darbu redzamu<\/h3>\n

    \nLiel\u0101kaj\u0101 da\u013c\u0101 uz\u0146\u0113mumu, ar kuriem es str\u0101d\u0101ju, ir \u013coti augsts nezin\u0101mas darba procents. Piem\u0113ram, kad viens darbinieks v\u0113r\u0161as pie otra un vienk\u0101r\u0161i l\u016bdz kaut ko izdar\u012bt. Liel\u0101s organiz\u0101cij\u0101s var b\u016bt l\u012bdz 60% nepl\u0101not\u0101 darba. Un pat l\u012bdz 40% darba nav dokument\u0113ts. Ja tas b\u016btu Boeing, es nekad vairs nelidotu ar vi\u0146u lidma\u0161\u012bnu. Ja tiek dokument\u0113ta tikai puse darba, nav skaidrs, vai \u0161is darbs tiek veikts pareizi vai n\u0113. Visi p\u0101r\u0113jie pa\u0146\u0113mieni izr\u0101d\u0101s bezj\u0113dz\u012bgi \u2014 nav j\u0113gas m\u0113\u0123in\u0101t kaut ko automatiz\u0113t, jo zin\u0101mie 50% var b\u016bt tie\u0161i visk\u0101rt\u012bg\u0101k\u0101 un prec\u012bz\u0101k\u0101 darba da\u013ca, kuras automatiz\u0101cija nenes\u012bs lielus rezult\u0101tus, bet viss \u0161ausm\u012bgais ir neredzamaj\u0101 pus\u0113. Bez dokument\u0101cijas ir neiesp\u0113jami atrast visda\u017e\u0101d\u0101k\u0101s nepiln\u012bbas un sl\u0113pto darbu, neatrast \u0161aur\u0101s vietas, tos pa\u0161us \u201cBrentus\u201d, par kuriem es jau run\u0101ju. Ir br\u012bni\u0161\u0137\u012bga gr\u0101mata no Dominikas De Grandis (Dominica DeGrandis) \u00abMaking Work Visible\u00bb<\/a><\/noindex>. T\u0101 identific\u0113 piecas da\u017e\u0101das \u00ablaika nopl\u016bdes\u00bb<\/b> (thieves of time):<\/p>\n