{"id":30548,"date":"2019-10-31T21:36:08","date_gmt":"2019-10-31T18:36:08","guid":{"rendered":"https:\/\/prohoster.info\/blog\/konfigurirovanie-spark-na-yarn\/"},"modified":"2019-10-31T21:36:08","modified_gmt":"2019-10-31T18:36:08","slug":"konfigurirovanie-spark-na-yarn","status":"publish","type":"post","link":"https:\/\/prohoster.info\/lv\/blog\/news\/konfigurirovanie-spark-na-yarn","title":{"rendered":"Spark konfigur\u0113\u0161ana uz YARN","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"<p>Sveiks, Habr! Vakar bija <noindex><a rel=\"nofollow\" href=\"https:\/\/www.facebook.com\/events\/1825957590998380\/\">mitaps, kas velt\u012bts Apache Spark<\/a><\/noindex>, no Rambler&#038;Co pui\u0161iem bija diezgan daudz jaut\u0101jumu no dal\u012bbniekiem par \u0161\u012b r\u012bka konfigur\u0113\u0161anu. Nol\u0113m\u0101m dal\u012bties ar savu pieredzi. T\u0113ma ir sare\u017e\u0123\u012bta \u2014 t\u0101p\u0113c ierosin\u0101m ar\u012b dal\u012bties ar pieredzi koment\u0101ros, varb\u016bt m\u0113s ar\u012b kaut ko nepareizi saprotam un izmantojam.<br \/>\n<noindex><a rel=\"nofollow\" name=\"habracut\"><\/a><\/noindex><br \/>\nNeliels ievads \u2014 k\u0101 m\u0113s izmantojam Spark. Mums ir tr\u012bs m\u0113ne\u0161u programma <noindex><a rel=\"nofollow\" href=\"http:\/\/newprolab.com\/ru\/bigdata?utm_source=habr&amp;utm_campaign=spark\">\u201eLielu datu speci\u0101lists\u201d<\/a><\/noindex>, un vis\u0101 otraj\u0101 modul\u012b m\u016bsu dal\u012bbnieki str\u0101d\u0101 ar \u0161o r\u012bku. T\u0101tad, m\u016bsu k\u0101 organizatoru uzdevums ir sagatavot klasteri, lai to izmantotu \u0161\u0101dos gad\u012bjumos.<\/p>\n<p>M\u016bsu izmanto\u0161anas \u012bpatn\u012bba ir t\u0101, ka cilv\u0113ku skaits, kas vienlaic\u012bgi str\u0101d\u0101 ar Spark, var b\u016bt vien\u0101ds ar visu grupu. Piem\u0113ram, semin\u0101r\u0101, kad visi vienlaic\u012bgi kaut ko izm\u0113\u0123ina un atk\u0101rto p\u0113c m\u016bsu pasniedz\u0113ja. Un tas ir diezgan daudz \u2014 reiz\u0113m l\u012bdz 40 cilv\u0113kiem. Dro\u0161i vien nav daudz uz\u0146\u0113mumu pasaul\u0113, kas sastopas ar \u0161\u0101du lieto\u0161anas scen\u0101riju.<\/p>\n<p>T\u0101l\u0101k past\u0101st\u012b\u0161u, k\u0101p\u0113c un k\u0101 m\u0113s izv\u0113l\u0113j\u0101mies noteiktus konfigur\u0101cijas parametrus.<\/p>\n<p>S\u0101ksim no pa\u0161a s\u0101kuma. Spark darbojas ar 3 iesp\u0113j\u0101m klaster\u012b: standalone, izmantojot Mesos un izmantojot YARN. M\u0113s nol\u0113m\u0101m izv\u0113l\u0113ties tre\u0161o variantu, jo tas mums \u0161\u0137ita lo\u0123isks. Mums jau ir Hadoop klasteris. M\u016bsu dal\u012bbnieki jau labi paz\u012bst t\u0101 arhitekt\u016bru. T\u0101p\u0113c izmantosim YARN.<\/p>\n<pre><code class=\"apache\">spark.master=yarn<\/code><\/pre>\n<p>\nT\u0101l\u0101k interesant\u0101k. Katram no \u0161iem 3 izvieto\u0161anas variantiem ir 2 izvieto\u0161anas varianti: client un cluster. \u0146emot v\u0113r\u0101 <noindex><a rel=\"nofollow\" href=\"http:\/\/spark.apache.org\/docs\/latest\/running-on-yarn.html\">dokument\u0101cij\u0101<\/a><\/noindex> un da\u017e\u0101d\u0101s saites internet\u0101, var secin\u0101t, ka client ir piem\u0113rots interakt\u012bvai darb\u012bbai \u2014 piem\u0113ram, caur Jupyter notebook, bet cluster vair\u0101k piem\u0113rots produkcijas risin\u0101jumiem. M\u016bsu gad\u012bjum\u0101 m\u016bs interes\u0113ja interakt\u012bv\u0101 darb\u012bba, t\u0101p\u0113c:<\/p>\n<pre><code class=\"apache\">spark.deploy-mode=client<\/code><\/pre>\n<p>\nPaties\u012bb\u0101 no \u0161\u012b br\u012b\u017ea Spark jau k\u0101d\u0101 m\u0113r\u0101 darbosies YARN, bet mums tas nebija pietiekami. T\u0101 k\u0101 mums ir programma par lieliem datiem, reiz\u0113m dal\u012bbniekiem tr\u016bka t\u0101, kas izn\u0101k ar vienm\u0113r\u012bgu resursu pie\u0161\u0137ir\u0161anu. Un \u0161eit m\u0113s atrad\u0101m interesantu lietu \u2014 dinamisko resursu pie\u0161\u0137ir\u0161anu. Ja \u012bsi, tad b\u016bt\u012bba ir sekojo\u0161a: ja jums ir smags uzdevums un klasteris ir br\u012bvs (piem\u0113ram, no r\u012bta), tad, izmantojot \u0161o opciju, Spark var jums pie\u0161\u0137irt papildu resursus. Nepiecie\u0161am\u012bba tiek apr\u0113\u0137in\u0101ta p\u0113c sare\u017e\u0123\u012btas formulas. P\u0101rl\u016bkot s\u012bkumus neiedzi\u013cin\u0101simies \u2014 t\u0101 darbojas diezgan labi.<\/p>\n<pre><code class=\"apache\">spark.dynamicAllocation.enabled=true<\/code><\/pre>\n<p>\nM\u0113s iestat\u012bj\u0101m \u0161o parametru, un, uzs\u0101kot Spark, tas uzsita un nes\u0101ka darboties. Pareizi, jo vajadz\u0113ja las\u012bt <noindex><a rel=\"nofollow\" href=\"http:\/\/spark.apache.org\/docs\/latest\/configuration.html\">dokumenti<\/a><\/noindex> uzman\u012bg\u0101k. Tur ir nor\u0101d\u012bts, ka, lai viss b\u016btu k\u0101rt\u012bb\u0101, j\u0101iesl\u0113dz v\u0113l viens papildu parametrs.<\/p>\n<pre><code class=\"apache\">spark.shuffle.service.enabled=true<\/code><\/pre>\n<p>\nK\u0101p\u0113c tas ir nepiecie\u0161ams? Kad m\u016bsu uzdevums vairs neprasa tik daudz resursu, tad Spark j\u0101atgrie\u017e tie kop\u0113j\u0101 basein\u0101. Visresursu pat\u0113r\u0113jo\u0161\u0101 f\u0101ze gandr\u012bz jebkur\u0101 MapReduce uzdevum\u0101 ir Shuffle f\u0101ze. \u0160is parametrs \u013cauj saglab\u0101t datus, kas veidojas \u0161aj\u0101 f\u0101z\u0113, un attiec\u012bgi atbr\u012bvot izpild\u012bt\u0101jus. Bet izpild\u012bt\u0101js ir process, kas uz darbinieka visu apr\u0113\u0137ina. Tam ir zin\u0101ms skaits procesoru kodolu un zin\u0101ms atmi\u0146as apjoms.<\/p>\n<p>Pievienoj\u0101m \u0161o parametru. Viss, \u0161\u0137iet, darboj\u0101s. Bija paman\u0101ms, ka dal\u012bbniekiem patie\u0161\u0101m tika izsniegti vair\u0101k resursu, kad tie bija vajadz\u012bgi. Bet rad\u0101s cita probl\u0113ma \u2014 k\u0101d\u0101 br\u012bd\u012b citi dal\u012bbnieki pamod\u0101s un ar\u012b v\u0113l\u0113j\u0101s izmantot Spark, bet tur viss bija aiz\u0146emts, un vi\u0146i nebija apmierin\u0101ti. Vi\u0146us var saprast. S\u0101k\u0101m skat\u012bties dokument\u0101cij\u0101. Tur izr\u0101d\u012bj\u0101s, ka ir v\u0113l vair\u0101ki parametri, ar kuru pal\u012bdz\u012bbu var ietekm\u0113t procesu. Piem\u0113ram, ja izpild\u012bt\u0101js ir gaid\u012b\u0161anas re\u017e\u012bm\u0101 \u2014 p\u0113c cik ilga laika taj\u0101 var at\u0146emt resursus?<\/p>\n<pre><code class=\"apache\">spark.dynamicAllocation.executorIdleTimeout=120s<\/code><\/pre>\n<p>\nM\u016bsu gad\u012bjum\u0101 \u2014 ja j\u016bsu executors divu min\u016b\u0161u laik\u0101 neko nedara, l\u016bdzu, atgrieziet tos kop\u0113j\u0101 basein\u0101. Tom\u0113r \u0161is parametrs ne vienm\u0113r bija pietiekams. Bija redzams, ka cilv\u0113ks jau sen neko nedara, bet resursi netiek atbr\u012bvoti. Izr\u0101d\u012bj\u0101s, ka ir v\u0113l viens \u012bpa\u0161s parametrs \u2014 p\u0113c cik ilga laika at\u0146emt executors, kuri satur ke\u0161atmi\u0146\u0101 glab\u0101tos datus. Noklus\u0113jum\u0101 \u0161is parametrs bija \u2014 bezgal\u012bba! M\u0113s to laboj\u0101m.<\/p>\n<pre><code class=\"apache\">spark.dynamicAllocation.cachedExecutorIdleTimeout=600s<\/code><\/pre>\n<p>\nProti, ja j\u016bsu executors neko nedara 5 min\u016btes, atdodiet tos kop\u0113j\u0101 basein\u0101. \u0160\u0101d\u0101 re\u017e\u012bm\u0101 resursu atbr\u012bvo\u0161anas un pie\u0161\u0137ir\u0161anas \u0101trums lielam lietot\u0101ju skaitam k\u013cuva apmierino\u0161s. Apmierin\u0101to skaits pieauga. Tom\u0113r m\u0113s nolem\u0101m iet t\u0101l\u0101k un ierobe\u017eot maksim\u0101lo executors skaitu vienai lietojumprogrammai \u2014 b\u016bt\u012bb\u0101 vienam programmas dal\u012bbniekam.<\/p>\n<pre><code class=\"apache\">spark.dynamicAllocation.maxExecutors=19<\/code><\/pre>\n<p>\nTagad, protams, par\u0101d\u012bj\u0101s neapmierin\u0101tie no citas puses \u2014 \u201cklasteris st\u0101v, bet man ir tikai 19 executors\u201d, bet kas j\u0101dara \u2014 nepiecie\u0161ams kaut k\u0101ds pareizs l\u012bdzsvars. Visiem laim\u012bgi padar\u012bt neizdosies.<\/p>\n<p>Un v\u0113l viena neliela st\u0101sta par m\u016bsu gad\u012bjuma specifiku. K\u0101du reizi uz praktisko nodarb\u012bbu kav\u0113j\u0101s da\u017ei cilv\u0113ki, un vi\u0146iem Spark k\u0101d\u0101 iemesla d\u0113\u013c nes\u0101k\u0101s. M\u0113s paskat\u012bj\u0101mies uz br\u012bvo resursu skaitu \u2014 it k\u0101 bija. Sparkam vajadz\u0113tu s\u0101kties. Par laimi, taj\u0101 br\u012bd\u012b dokument\u0101cija jau kaut kur bija ierakst\u012bjusies atmi\u0146\u0101, un m\u0113s atcer\u0113j\u0101mies, ka, palai\u017eot Spark, tas mekl\u0113 portu, kur\u0101 s\u0101kt. Ja pirmais ports no diapazona ir aiz\u0146emts, tas p\u0101riet uz n\u0101kamo. Ja tas ir br\u012bvs, tas to izmanto. Un ir parametrs, kas nor\u0101da maksim\u0101lo m\u0113\u0123in\u0101jumu skaitu \u0161im. Noklus\u0113jum\u0101 \u2014 tas ir 16. \u0160is skaitlis ir maz\u0101ks par cilv\u0113ku skaitu m\u016bsu grup\u0101 nodarb\u012bb\u0101. L\u012bdz ar to p\u0113c 16 m\u0113\u0123in\u0101jumiem Spark atsaka un saka, ka nevar s\u0101kties. M\u0113s laboj\u0101m \u0161o parametru.<\/p>\n<pre><code class=\"apache\">spark.port.maxRetries=50<\/code><\/pre>\n<p>\nT\u0101l\u0101k past\u0101st\u012b\u0161u par da\u017eiem uzst\u0101d\u012bjumiem, kas nav \u013coti saist\u012bti ar m\u016bsu gad\u012bjuma specifiku.<\/p>\n<p>Lai Spark \u0101tr\u0101k iel\u0101d\u0113tos, ir ieteikums arhiv\u0113t jars mapes, kas atrodas SPARK_HOME m\u0101jas direktorij\u0101, un novietot to HDFS. Tad tas net\u0113r\u0113s laiku \u0161\u012b jar failu lejupiel\u0101dei darba ma\u0161\u012bn\u0101s.<\/p>\n<pre><code class=\"apache\">spark.yarn.archive=hdfs:\/\/\/tmp\/spark-archive.zip<\/code><\/pre>\n<p>\nLai uzlabotu veiktsp\u0113ju, ieteicams izmantot kriyo k\u0101 s\u0113rija. Tas ir optimiz\u0113t\u0101ks nek\u0101 noklus\u0113juma variants.<\/p>\n<pre><code class=\"apache\">spark.serializer=org.apache.spark.serializer.KryoSerializer<\/code><\/pre>\n<p>\nIr v\u0113l sen zin\u0101ma probl\u0113ma ar Spark, ka tas bie\u017ei iznine sapr\u0101tu. Bie\u017ei tas notiek br\u012bd\u012b, kad darbinieki visu apr\u0113\u0137ina un nos\u016bta rezult\u0101tu uz draiveri. M\u0113s palielin\u0101j\u0101m \u0161o parametru. Noklus\u0113juma v\u0113rt\u012bba ir 1GB, bet m\u0113s to izdar\u012bj\u0101m \u2014 3.<\/p>\n<pre><code class=\"apache\">spark.driver.maxResultSize=3072<\/code><\/pre>\n<p>\nUn p\u0113d\u0113jais, k\u0101 deserts. K\u0101 atjaunin\u0101t Spark uz 2.1 versiju HortonWorks sadal\u012bjum\u0101 \u2014 HDP 2.5.3.0. \u0160\u012b HDP versija satur iepriek\u0161 instal\u0113tu 2.0 versiju, bet k\u0101du reizi nol\u0113m\u0101m, ka Spark att\u012bst\u0101s diezgan akt\u012bvi, un katra jaun\u0101 versija nov\u0113r\u0161 kaut k\u0101das k\u013c\u016bdas un pied\u0101v\u0101 papildu iesp\u0113jas, tostarp python API, t\u0101d\u0113\u013c nol\u0113m\u0101m veikt atjaunin\u0101jumu.<\/p>\n<p>Lejupiel\u0101d\u0113j\u0101m versiju no ofici\u0101l\u0101s m\u0101jaslapas, kas paredz\u0113ta Hadoop 2.7. Izsme\u013cot, ievietoj\u0101m HDP map\u0113. Izveidoj\u0101m nepiecie\u0161amos simboliskos saites. Uzs\u0101kam \u2014 nes\u0101kas. Par\u0101d\u0101s \u013coti neskaidra k\u013c\u016bda.<\/p>\n<pre><code class=\"apache\">java.lang.NoClassDefFoundError: com\/sun\/jersey\/api\/client\/config\/ClientConfig<\/code><\/pre>\n<p>\nP\u0101rl\u016bkojot, noskaidroj\u0101m, ka Spark nol\u0113ma negaid\u012bt, kam\u0113r Hadoop izn\u0101ks, un nol\u0113ma izmantot jauno jersey versiju. Vi\u0146i pa\u0161i tur viens ar otru str\u012bdas par \u0161o t\u0113mu JIRA. Risin\u0101jums bija \u2014 lejupiel\u0101d\u0113t <noindex><a rel=\"nofollow\" href=\"https:\/\/mvnrepository.com\/artifact\/com.sun.jersey\/jersey-bundle\/1.17.1\">jersey versija 1.17.1<\/a><\/noindex>. Ievietot to j\u0101r\u0101m SPARK_HOME map\u0113, atkal izveidot zip un aug\u0161upiel\u0101d\u0113t uz HDFS.<\/p>\n<p>M\u0113s \u0161o k\u013c\u016bdu apg\u0101j\u0101m, bet rad\u0101s jauna un diezgan miglaina.<\/p>\n<pre><code class=\"apache\">org.apache.spark.SparkException: Yarn application has already ended! It might have been killed or unable to launch application master<\/code><\/pre>\n<p>\nTaj\u0101 pa\u0161\u0101 laik\u0101 m\u0113\u0123in\u0101m palaist 2.0 versiju \u2014 viss k\u0101rt\u012bb\u0101. M\u0113\u0123ini uzmin\u0113t, kas par iemeslu. M\u0113s iegrie\u017eamies \u0161\u012b pieteikuma \u017eurn\u0101los un redzam kaut ko t\u0101du:<\/p>\n<pre><code class=\"apache\">\/usr\/hdp\/${hdp.version}\/hadoop\/lib\/hadoop-lzo-0.6.0.${hdp.version}.jar<\/code><\/pre>\n<p>\nKopum\u0101 k\u0101du iemeslu d\u0113\u013c hdp.version netika atrisin\u0101ta. P\u0101rl\u016bkojot, atrad\u0101m risin\u0101jumu. J\u0101dodas uz Ambari un j\u0101pieprasa YARN iestat\u012bjumos pievienot parametru custom yarn-site:<\/p>\n<pre><code class=\"apache\">hdp.version=2.5.3.0-37<\/code><\/pre>\n<p>\n\u0160\u012b ma\u0123ija pal\u012bdz\u0113ja, un Spark pac\u0113l\u0101s. M\u0113s p\u0101rbaud\u012bj\u0101m vair\u0101kus m\u016bsu jupyter piez\u012bmjdatorus. Viss darbojas. M\u0113s esam gatavi pirmajai nodarb\u012bbai par Spark sestdien (jau r\u012bt)!<\/p>\n<p><b>UPD<\/b>. Nodarb\u012bbas laik\u0101 atkl\u0101j\u0101s v\u0113l viena probl\u0113ma. K\u0101d\u0101 br\u012bd\u012b YARN p\u0101rst\u0101ja pie\u0161\u0137irt konteinerus Spark. YARN vajadz\u0113ja piel\u0101got parametrus, kas noklus\u0113jum\u0101 bija 0.2:<\/p>\n<pre><code class=\"apache\">yarn.scheduler.capacity.maximum-am-resource-percent=0.8<\/code><\/pre>\n<p>\nT\u0101tad tikai 20% resursu piedal\u012bj\u0101s resursu sadal\u0113. Mainot iestat\u012bjumus, p\u0101rstart\u0113j\u0101m YARN. Probl\u0113ma tika atrisin\u0101ta, un p\u0101r\u0113jie dal\u012bbnieki ar\u012b var\u0113ja palaist spark kontekstu.<br \/>\n<br \/>Avots: <a content=\"nofollow\" rel=\"nofollow\" href=\"https:\/\/habr.com\/ru\/company\/newprolab\/blog\/327556\/\">habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"<p>\u0425\u0430\u0431\u0440, \u043f\u0440\u0438\u0432\u0435\u0442! \u0412\u0447\u0435\u0440\u0430 \u043d\u0430 \u043c\u0438\u0442\u0430\u043f\u0435, \u043f\u043e\u0441\u0432\u044f\u0449\u0435\u043d\u043d\u043e\u043c Apache Spark, \u043e\u0442 \u0440\u0435\u0431\u044f\u0442 \u0438\u0437 Rambler&#038;Co, \u0431\u044b\u043b\u043e \u0434\u043e\u0432\u043e\u043b\u044c\u043d\u043e \u043c\u043d\u043e\u0433\u043e \u0432\u043e\u043f\u0440\u043e\u0441\u043e\u0432 \u043e\u0442 \u0443\u0447\u0430\u0441\u0442\u043d\u0438\u043a\u043e\u0432, \u0441\u0432\u044f\u0437\u0430\u043d\u043d\u044b\u0445 \u0441 \u043a\u043e\u043d\u0444\u0438\u0433\u0443\u0440\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0435\u043c \u044d\u0442\u043e\u0433\u043e \u0438\u043d\u0441\u0442\u0440\u0443\u043c\u0435\u043d\u0442\u0430. \u0420\u0435\u0448\u0438\u043b\u0438 \u043f\u043e \u0435\u0433\u043e \u0441\u043b\u0435\u0434\u0430\u043c \u043f\u043e\u0434\u0435\u043b\u0438\u0442\u044c\u0441\u044f \u0441\u0432\u043e\u0438\u043c \u043e\u043f\u044b\u0442\u043e\u043c. \u0422\u0435\u043c\u0430 \u043d\u0435\u043f\u0440\u043e\u0441\u0442\u0430\u044f \u2014 \u043f\u043e\u044d\u0442\u043e\u043c\u0443 \u043f\u0440\u0435\u0434\u043b\u0430\u0433\u0430\u0435\u043c \u0434\u0435\u043b\u0438\u0442\u044c\u0441\u044f \u043e\u043f\u044b\u0442\u043e\u043c \u0442\u043e\u0436\u0435 \u0432 \u043a\u043e\u043c\u043c\u0435\u043d\u0442\u0430\u0440\u0438\u044f\u0445, \u043c\u043e\u0436\u0435\u0442 \u0431\u044b\u0442\u044c, \u043c\u044b \u0442\u043e\u0436\u0435 \u0447\u0442\u043e-\u0442\u043e \u043d\u0435 \u0442\u0430\u043a \u043f\u043e\u043d\u0438\u043c\u0430\u0435\u043c \u0438 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u043c. \u041d\u0435\u0431\u043e\u043b\u044c\u0448\u0430\u044f \u0432\u0432\u043e\u0434\u043d\u0430\u044f \u2014 \u043a\u0430\u043a \u043c\u044b [&hellip;]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-30548","post","type-post","status-publish","format-standard","hentry","category-news"],"aioseo_notices":[],"aioseo_head":"\n\t\t<!-- All in One SEO 5.0.5 - aioseo.com -->\n\t<meta name=\"description\" content=\"\u0425\u0430\u0431\u0440, \u043f\u0440\u0438\u0432\u0435\u0442!\" \/>\n\t<meta name=\"robots\" content=\"max-image-preview:large\" \/>\n\t<meta name=\"author\" content=\"Yuri Gagarin\"\/>\n\t<link rel=\"canonical\" href=\"https:\/\/prohoster.info\/lv\/blog\/news\/konfigurirovanie-spark-na-yarn\" \/>\n\t\t<meta name=\"generator\" content=\"All in One SEO (AIOSEO) 5.0.5\" \/>\n\t\t<meta property=\"og:locale\" content=\"lv_LV\" \/>\n\t\t<meta property=\"og:site_name\" content=\"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b\" \/>\n\t\t<meta property=\"og:type\" content=\"article\" \/>\n\t\t<meta property=\"og:title\" content=\"\ud83e\udd47\u041a\u043e\u043d\u0444\u0438\u0433\u0443\u0440\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0435 Spark \u043d\u0430 YARN | ProHoster\" \/>\n\t\t<meta property=\"og:description\" content=\"\u0425\u0430\u0431\u0440, \u043f\u0440\u0438\u0432\u0435\u0442!\" \/>\n\t\t<meta property=\"og:url\" content=\"https:\/\/prohoster.info\/lv\/blog\/news\/konfigurirovanie-spark-na-yarn\" \/>\n\t\t<meta property=\"og:image\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:secure_url\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:width\" content=\"350\" \/>\n\t\t<meta property=\"og:image:height\" content=\"350\" \/>\n\t\t<meta property=\"article:published_time\" content=\"2019-10-31T18:36:08+00:00\" \/>\n\t\t<meta property=\"article:modified_time\" content=\"2019-10-31T18:36:08+00:00\" \/>\n\t\t<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<meta property=\"article:author\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<!-- All in One SEO -->\n\n","aioseo_head_json":{"title":"\ud83e\udd47Spark konfigur\u0113\u0161ana uz YARN | ProHoster","description":"Habr, sveiki!","canonical_url":"https:\/\/prohoster.info\/lv\/blog\/news\/konfigurirovanie-spark-na-yarn","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"lv_LV","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd47\u041a\u043e\u043d\u0444\u0438\u0433\u0443\u0440\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0435 Spark \u043d\u0430 YARN | ProHoster","og:description":"\u0425\u0430\u0431\u0440, \u043f\u0440\u0438\u0432\u0435\u0442!","og:url":"https:\/\/prohoster.info\/lv\/blog\/news\/konfigurirovanie-spark-na-yarn","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2019-10-31T18:36:08+00:00","article:modified_time":"2019-10-31T18:36:08+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"30548","title":null,"description":null,"keywords":null,"keyphrases":null,"primary_term":null,"canonical_url":null,"og_title":null,"og_description":null,"og_object_type":"default","og_image_type":"default","og_image_url":null,"og_image_width":null,"og_image_height":null,"og_image_custom_url":null,"og_image_custom_fields":null,"og_video":null,"og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":false,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"","isEnabled":true},"graphs":[]},"schema_type":null,"schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":null,"robots_max_videopreview":null,"robots_max_imagepreview":"large","priority":null,"frequency":null,"local_seo":null,"seo_analyzer_scan_date":"2026-01-21 01:45:22","breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-03-01 03:33:26","updated":"2026-01-21 01:45:22","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/posts\/30548","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/comments?post=30548"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/posts\/30548\/revisions"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/media?parent=30548"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/categories?post=30548"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/lv\/wp-json\/wp\/v2\/tags?post=30548"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}