{"id":191769,"date":"2023-03-08T18:39:00","date_gmt":"2023-03-08T18:39:00","guid":{"rendered":"https:\/\/datascientest.com\/?p=191769"},"modified":"2023-11-09T13:53:53","modified_gmt":"2023-11-09T13:53:53","slug":"apache-flink-tout-savoir","status":"publish","type":"post","link":"https:\/\/datascientest.com\/apache-flink-tout-savoir","title":{"rendered":"Apache Flink : D\u00e9finition, structure et limites du framework"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"191769\" class=\"elementor elementor-191769\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-3a37298 e-flex e-con-boxed e-con\" data-id=\"3a37298\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-6b9e08d elementor-widget elementor-widget-heading\" data-id=\"6b9e08d\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-heading-title{padding:0;margin:0;line-height:1}.elementor-widget-heading .elementor-heading-title[class*=elementor-size-]>a{color:inherit;font-size:inherit;line-height:inherit}.elementor-widget-heading .elementor-heading-title.elementor-size-small{font-size:15px}.elementor-widget-heading .elementor-heading-title.elementor-size-medium{font-size:19px}.elementor-widget-heading .elementor-heading-title.elementor-size-large{font-size:29px}.elementor-widget-heading .elementor-heading-title.elementor-size-xl{font-size:39px}.elementor-widget-heading .elementor-heading-title.elementor-size-xxl{font-size:59px}<\/style><h3 class=\"elementor-heading-title elementor-size-default\">Les origines de Apache Flink<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e51bf12 elementor-widget elementor-widget-text-editor\" data-id=\"e51bf12\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-text-editor.elementor-drop-cap-view-stacked .elementor-drop-cap{background-color:#69727d;color:#fff}.elementor-widget-text-editor.elementor-drop-cap-view-framed .elementor-drop-cap{color:#69727d;border:3px solid;background-color:transparent}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap{margin-top:8px}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap-letter{width:1em;height:1em}.elementor-widget-text-editor .elementor-drop-cap{float:left;text-align:center;line-height:1;font-size:50px}.elementor-widget-text-editor .elementor-drop-cap-letter{display:inline-block}<\/style>\t\t\t\t<p><span style=\"font-weight: 400;\">Initialement d\u00e9velopp\u00e9 \u00e0 l\u2019universit\u00e9 technique de Berlin, ses premi\u00e8res versions ont \u00e9t\u00e9 publi\u00e9es en 2011 et avaient pour but de <strong>r\u00e9pondre \u00e0 des probl\u00e9matiques complexes dans le traitement de donn\u00e9es dans un environnement distribu\u00e9 en temps r\u00e9el<\/strong>. Flink est devenu au fil des ann\u00e9es une r\u00e9f\u00e9rence pour un grand nombre d\u2019entreprises, jusqu\u2019\u00e0 devenir un des framework open-source les plus populaires. C\u2019est en 2014 que Flink a \u00e9t\u00e9 accept\u00e9 en tant que projet Apache Incubator, puis en 2015 est devenu un projet Apache Top-Level. Depuis lors, Flink n\u2019a eu de cesse de s\u2019am\u00e9liorer, fier d\u2019<strong>une communaut\u00e9 de d\u00e9veloppeurs et d\u2019utilisateurs tr\u00e8s active<\/strong>.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-319b0f9 e-flex e-con-boxed e-con\" data-id=\"319b0f9\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-8a47117 elementor-widget elementor-widget-heading\" data-id=\"8a47117\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Composition de l'\u00e9cosyst\u00e8me Flink<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-421d54a elementor-widget elementor-widget-text-editor\" data-id=\"421d54a\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">L\u2019\u00e9cosyst\u00e8me de Apache Flink est compos\u00e9 de plusieurs couches et niveaux d\u2019abstraction, comme illustr\u00e9 dans la figure suivante :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-84660b4 elementor-widget elementor-widget-image\" data-id=\"84660b4\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-image{text-align:center}.elementor-widget-image a{display:inline-block}.elementor-widget-image a img[src$=\".svg\"]{width:48px}.elementor-widget-image img{vertical-align:middle;display:inline-block}<\/style>\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"554\" height=\"416\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image4-3.png\" class=\"attachment-full size-full wp-image-191772\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image4-3.png 554w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image4-3-300x225.png 300w\" sizes=\"(max-width: 554px) 100vw, 554px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-aedaedf elementor-widget elementor-widget-text-editor\" data-id=\"aedaedf\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<ul><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Storage <\/b><span style=\"font-weight: 400;\">(stockage) : Flink dispose de plusieurs possibilit\u00e9s pour lire \/ \u00e9crire les donn\u00e9es, telles que HDFS (<a href=\"https:\/\/datascientest.com\/hadoop\">Hadoop<\/a>), S3, en local, Kafka, et bien d\u2019autres<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Deploy <\/b><span style=\"font-weight: 400;\">(d\u00e9ploiement) : Flink peut se d\u00e9ployer en mode local, sur des clusters ou dans le <a href=\"https:\/\/datascientest.com\/comment-devenir-cloud-expert\">cloud<\/a><\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Kernel <\/b><span style=\"font-weight: 400;\">(noyau) : Il s\u2019agit ici de la couche d\u2019ex\u00e9cution, qui donne la tol\u00e9rance aux pannes, les diff\u00e9rents calculs distribu\u00e9s, etc.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>API\u2019s &amp; Libraries<\/b> (<a href=\"https:\/\/datascientest.com\/api\">API <\/a>et biblioth\u00e8ques) : Il s\u2019agit de la couche de plus haut niveau de l\u2019\u00e9cosyst\u00e8me Flink. On y retrouve l\u2019API Datastream, en charge du traitement des flux, l\u2019API Dataset, en charge du traitement par lot (batch processing), et d\u2019autres biblioth\u00e8ques comme Flink ML (Machine Learning), Gelly (graph processing) et Table (pour le <a href=\"https:\/\/datascientest.com\/ouvrir-un-fichier-sql-tutoriel\">SQL<\/a>).<\/li><\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9580f94 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"9580f94\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formations-data\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Se former \u00e0 l'\u00e9cosyst\u00e8me Flink<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-28903f1 e-flex e-con-boxed e-con\" data-id=\"28903f1\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-8875ee5 elementor-widget elementor-widget-heading\" data-id=\"8875ee5\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Comment est structur\u00e9 l'architecture de Flink ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bf80f25 elementor-widget elementor-widget-text-editor\" data-id=\"bf80f25\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Flink est un moteur de traitement distribu\u00e9 pour des calculs avec \u00e9tat sur des flux de donn\u00e9es <\/span><b>born\u00e9s <\/b><span style=\"font-weight: 400;\">ou <\/span><b>illimit\u00e9s<\/b><span style=\"font-weight: 400;\">. Il a \u00e9t\u00e9 con\u00e7u pour fonctionner dans tous les environnements de cluster courants, et pour \u00eatre en mesure d\u2019effectuer des calculs \u00e0 une vitesse tr\u00e8s \u00e9lev\u00e9e, et \u00e0 n\u2019importe quelle \u00e9chelle.<\/span><\/p><p><span style=\"font-weight: 400;\">Tout type de donn\u00e9es est produit sous forme de flux, qu\u2019il s&rsquo;agisse de relev\u00e9s provenant de capteurs (<\/span><b>IoT<\/b><span style=\"font-weight: 400;\">), de journaux d&rsquo;\u00e9v\u00e9nements (<\/span><b>logs<\/b><span style=\"font-weight: 400;\">) ou m\u00eame d\u2019activit\u00e9s utilisateurs sur des sites web.<\/span><\/p><p><span style=\"font-weight: 400;\">Comme nous venons de le voir, les flux trait\u00e9s peuvent \u00eatre soit <\/span><b>born\u00e9s<\/b><span style=\"font-weight: 400;\">, soit <\/span><b>illimit\u00e9s<\/b><span style=\"font-weight: 400;\">, mais que cela signifie-t-il exactement ?\u00a0<\/span><\/p><ul><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Les <\/span><b>flux illimit\u00e9s<\/b><span style=\"font-weight: 400;\"> ont un d\u00e9but, mais pas de fin d\u00e9finie, ce qui signifie donc qu\u2019ils fournissent les donn\u00e9es telles qu\u2019elles sont g\u00e9n\u00e9r\u00e9es. Ils doivent \u00eatre trait\u00e9s en continu, c\u2019est-\u00e0-dire imm\u00e9diatement apr\u00e8s ingestion. De ce fait, il est impossible d\u2019attendre que tout le flux soit arriv\u00e9 pour le traiter, mais en revanche, il sera n\u00e9cessaire de traiter ces flux dans l\u2019ordre dans lequel ils arrivent.\u00a0<\/span><ul><li style=\"font-weight: 400;\" aria-level=\"2\"><b>Le contr\u00f4le pr\u00e9cis du temps et de l\u2019\u00e9tat permet \u00e0 Flink d\u2019ex\u00e9cuter n\u2019importe quel type d\u2019application sur ces flux.<\/b><\/li><\/ul><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Les <\/span><b>flux limit\u00e9s<\/b><span style=\"font-weight: 400;\">, ou <\/span><b>born\u00e9s<\/b><span style=\"font-weight: 400;\">, ont quant \u00e0 eux un d\u00e9but et une fin d\u00e9finis. Contrairement aux flux illimit\u00e9s, ils peuvent \u00eatre trait\u00e9s apr\u00e8s avoir ing\u00e9r\u00e9 l\u2019int\u00e9gralit\u00e9 des donn\u00e9es pour effectuer des calculs. Leurs traitements ordonn\u00e9s ne sont par cons\u00e9quent pas n\u00e9cessaires. Les traitements de ces flux sont \u00e9galement appel\u00e9s <\/span><b>traitement par lots <\/b><span style=\"font-weight: 400;\">ou <\/span><b>batch-processing<\/b><span style=\"font-weight: 400;\">.\u00a0<\/span><ul><li style=\"font-weight: 400;\" aria-level=\"2\"><b>Ces flux sont trait\u00e9s en interne par des algorithmes et des calculs sp\u00e9cialement con\u00e7us pour les datasets de taille fixe.<\/b><\/li><\/ul><\/li><\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bbf099b elementor-widget elementor-widget-image\" data-id=\"bbf099b\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"1999\" height=\"474\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image1-3.png\" class=\"attachment-full size-full wp-image-191776\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image1-3.png 1999w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image1-3-300x71.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image1-3-1024x243.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image1-3-768x182.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image1-3-1536x364.png 1536w\" sizes=\"(max-width: 1999px) 100vw, 1999px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4f5404e elementor-widget elementor-widget-text-editor\" data-id=\"4f5404e\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">On comprend alors facilement que Flink propose d\u2019excellente performance dans le traitement d\u2019ensemble de donn\u00e9es born\u00e9s ou illimit\u00e9s.\u00a0<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-e0dad98 e-flex e-con-boxed e-con\" data-id=\"e0dad98\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-ac021ff elementor-widget elementor-widget-heading\" data-id=\"ac021ff\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Quelles sont les applications li\u00e9es \u00e0 Flink gr\u00e2ce aux API\u2019s ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-222c159 elementor-widget elementor-widget-text-editor\" data-id=\"222c159\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Apache Flink propose un ensemble d\u2019<a href=\"https:\/\/datascientest.com\/construire-des-api-premiere-partie\">API<\/a> tr\u00e8s riches permettant de faire des transformations \u00e0 la fois sur des donn\u00e9es par lot (<\/span><b>batch<\/b><span style=\"font-weight: 400;\">) ou en temps-r\u00e9el (<\/span><b>streaming<\/b><span style=\"font-weight: 400;\">). Ces transformations sont effectu\u00e9es sur les donn\u00e9es distribu\u00e9es et permettent aux d\u00e9veloppeurs de construire une application qui r\u00e9pondra \u00e0 leur besoin. Voici les principales APIs propos\u00e9es par Flink :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-46ca356 elementor-widget elementor-widget-heading\" data-id=\"46ca356\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-medium\">L\u2019API Datastream :<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-a29b72c elementor-widget elementor-widget-text-editor\" data-id=\"a29b72c\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Cette API est la principale interface de programmation pour la cr\u00e9ation de flux de donn\u00e9es (<\/span><b>stream processing<\/b><span style=\"font-weight: 400;\">).<\/span><\/p><p><span style=\"font-weight: 400;\">Elle fournit une abstraction de haut niveau pour leur manipulation, et offre une prise en charge native pour la r\u00e9partition (partitioning) des donn\u00e9es pour le traitement en parall\u00e8le.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-a91dbdf elementor-widget elementor-widget-heading\" data-id=\"a91dbdf\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-medium\">L\u2019API Dataset :<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-af70dd9 elementor-widget elementor-widget-text-editor\" data-id=\"af70dd9\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Cette API est quant \u00e0 elle une interface de programmation pour le traitement de donn\u00e9es par lots (<\/span><b>batch processing<\/b><span style=\"font-weight: 400;\">). Elle fournit \u00e9galement une prise en charge native pour le traitement en parall\u00e8le \u00e0 grande \u00e9chelle.<\/span><\/p><p><span style=\"font-weight: 400;\">Elle est bas\u00e9e sur le mod\u00e8le de <\/span><a href=\"https:\/\/datascientest.com\/mapreduce\">MapReduce<\/a> <span style=\"font-weight: 400;\">mais offrant des possibilit\u00e9s plus avanc\u00e9es, comme la gestion de l\u2019\u00e9tat distribu\u00e9, la gestion des erreurs et une optimisation automatique.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-42246af elementor-widget elementor-widget-heading\" data-id=\"42246af\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-medium\">L\u2019API Table :<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1196958 elementor-widget elementor-widget-text-editor\" data-id=\"1196958\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Cette API est une interface de programmation pour le traitement de <\/span><b>donn\u00e9es relationnelles<\/b><span style=\"font-weight: 400;\">. Elle permet aux d\u00e9veloppeurs de cr\u00e9er des programmes de traitement de donn\u00e9es bas\u00e9s sur des tables et des requ\u00eates SQL, tout en profitant des avantages offerts par Flink.<\/span><\/p><p><span style=\"font-weight: 400;\">Elle est compatible avec les autres APIs de Flink, permettant ainsi de combiner facilement les diff\u00e9rentes interfaces pour r\u00e9pondre \u00e0 des besoins de traitement de donn\u00e9es sp\u00e9cifiques.<\/span><\/p><p><span style=\"font-weight: 400;\">Les programmes cr\u00e9\u00e9s avec ces API sont facilement <\/span><b>\u00e9volutifs <\/b><span style=\"font-weight: 400;\">et r<\/span><b>\u00e9silients aux pannes<\/b><span style=\"font-weight: 400;\"> gr\u00e2ce \u00e0 l\u2019architecture distribu\u00e9e de Flink. Ils peuvent \u00eatre ex\u00e9cut\u00e9s sur des clusters de traitement \u00e0 grande \u00e9chelle, garantissant ainsi des performances \u00e9lev\u00e9es et une haute disponibilit\u00e9.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6f22d2e elementor-blockquote--skin-border elementor-blockquote--button-color-official elementor-widget elementor-widget-blockquote\" data-id=\"6f22d2e\" data-element_type=\"widget\" data-widget_type=\"blockquote.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor-pro - v3.15.0 - 09-08-2023 *\/\n@charset \"UTF-8\";.entry-content blockquote.elementor-blockquote:not(.alignright):not(.alignleft),.entry-summary blockquote.elementor-blockquote{margin-right:0;margin-left:0}.elementor-widget-blockquote blockquote{margin:0;padding:0;outline:0;font-size:100%;vertical-align:baseline;background:transparent;quotes:none;border:0;font-style:normal;color:#3f444b}.elementor-widget-blockquote blockquote .e-q-footer:after,.elementor-widget-blockquote blockquote .e-q-footer:before,.elementor-widget-blockquote blockquote:after,.elementor-widget-blockquote blockquote:before,.elementor-widget-blockquote blockquote cite:after,.elementor-widget-blockquote blockquote cite:before{content:none}.elementor-blockquote{transition:.3s}.elementor-blockquote__author,.elementor-blockquote__content{margin-bottom:0;font-style:normal}.elementor-blockquote__author{font-weight:700}.elementor-blockquote .e-q-footer{margin-top:12px;display:flex;justify-content:space-between}.elementor-blockquote__tweet-button{display:flex;transition:.3s;color:#1da1f2;align-self:flex-end;line-height:1;position:relative;width:-moz-max-content;width:max-content}.elementor-blockquote__tweet-button:hover{color:#0967a0}.elementor-blockquote__tweet-button span{font-weight:600}.elementor-blockquote__tweet-button i,.elementor-blockquote__tweet-button span{vertical-align:middle}.elementor-blockquote__tweet-button i+span,.elementor-blockquote__tweet-button svg+span{margin-left:.5em}.elementor-blockquote__tweet-button svg{fill:#1da1f2;height:1em;width:1em}.elementor-blockquote__tweet-label{white-space:pre-wrap}.elementor-blockquote--button-skin-bubble .elementor-blockquote__tweet-button,.elementor-blockquote--button-skin-classic .elementor-blockquote__tweet-button{padding:.7em 1.2em;border-radius:100em;background-color:#1da1f2;color:#fff;font-size:15px}.elementor-blockquote--button-skin-bubble .elementor-blockquote__tweet-button:hover,.elementor-blockquote--button-skin-classic .elementor-blockquote__tweet-button:hover{background-color:#0967a0;color:#fff}.elementor-blockquote--button-skin-bubble .elementor-blockquote__tweet-button:hover:before,.elementor-blockquote--button-skin-classic .elementor-blockquote__tweet-button:hover:before{border-right-color:#0967a0}.elementor-blockquote--button-skin-bubble .elementor-blockquote__tweet-button svg,.elementor-blockquote--button-skin-classic .elementor-blockquote__tweet-button svg{fill:#fff;height:1em;width:1em}.elementor-blockquote--button-skin-bubble.elementor-blockquote--button-view-icon .elementor-blockquote__tweet-button,.elementor-blockquote--button-skin-classic.elementor-blockquote--button-view-icon .elementor-blockquote__tweet-button{padding:0;width:2em;height:2em}.elementor-blockquote--button-skin-bubble.elementor-blockquote--button-view-icon .elementor-blockquote__tweet-button i,.elementor-blockquote--button-skin-classic.elementor-blockquote--button-view-icon .elementor-blockquote__tweet-button i{position:absolute;left:50%;top:50%;transform:translate(-50%,-50%)}.elementor-blockquote--button-skin-bubble .elementor-blockquote__tweet-button:before{content:\"\";border:.5em solid transparent;border-right-color:#1da1f2;position:absolute;left:-.8em;top:50%;transform:translateY(-50%) scaleY(.65);transition:.3s}.elementor-blockquote--button-skin-bubble.elementor-blockquote--align-left .elementor-blockquote__tweet-button:before{right:auto;left:-.8em;border-right-color:#1da1f2;border-left-color:transparent}.elementor-blockquote--button-skin-bubble.elementor-blockquote--align-left .elementor-blockquote__tweet-button:hover:before{border-right-color:#0967a0}.elementor-blockquote--button-skin-bubble.elementor-blockquote--align-right .elementor-blockquote__tweet-button:before{left:auto;right:-.8em;border-right-color:transparent;border-left-color:#1da1f2}.elementor-blockquote--button-skin-bubble.elementor-blockquote--align-right .elementor-blockquote__tweet-button:hover:before{border-left-color:#0967a0}.elementor-blockquote--skin-boxed .elementor-blockquote{background-color:#f9fafa;padding:30px}.elementor-blockquote--skin-border .elementor-blockquote{border-color:#f9fafa;border-left:7px #f9fafa;border-style:solid;padding-left:20px}.elementor-blockquote--skin-quotation .elementor-blockquote:before{content:\"\u201c\";font-size:100px;color:#f9fafa;font-family:Times New Roman,Times,serif;font-weight:900;line-height:1;display:block;height:.6em}.elementor-blockquote--skin-quotation .elementor-blockquote__content{margin-top:15px}.elementor-blockquote--align-left .elementor-blockquote__content{text-align:left}.elementor-blockquote--align-left .elementor-blockquote .e-q-footer{flex-direction:row}.elementor-blockquote--align-right .elementor-blockquote__content{text-align:right}.elementor-blockquote--align-right .elementor-blockquote .e-q-footer{flex-direction:row-reverse}.elementor-blockquote--align-center .elementor-blockquote{text-align:center}.elementor-blockquote--align-center .elementor-blockquote .e-q-footer,.elementor-blockquote--align-center .elementor-blockquote__author{display:block}.elementor-blockquote--align-center .elementor-blockquote__tweet-button{margin-right:auto;margin-left:auto}<\/style>\t\t<blockquote class=\"elementor-blockquote\">\n\t\t\t<p class=\"elementor-blockquote__content\">\n\t\t\t\t<u>\u00c0 noter<\/u> : Il existe \u00e9galement les API\u2019s <b>FlinkML<\/b> pour le Machine Learning, <b>Gelly<\/b> pour les bases orient\u00e9es graph ou encore <b>CEP<\/b> pour les traitements complexes (Complex Event Processing).\t\t\t<\/p>\n\t\t\t\t\t<\/blockquote>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b520a21 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"b520a21\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formations-data\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Apprendre \u00e0 concevoir des APIs<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-9075619 e-flex e-con-boxed e-con\" data-id=\"9075619\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-b0a1128 elementor-widget elementor-widget-heading\" data-id=\"b0a1128\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Fonctionnalit\u00e9s cl\u00e9s de Flink<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-485e616 elementor-widget elementor-widget-text-editor\" data-id=\"485e616\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Apache Flink dispose des notions cl\u00e9s \u00e9num\u00e9r\u00e9es ci-apr\u00e8s :\u00a0<\/span><\/p><ul><li style=\"font-weight: 400;\" aria-level=\"1\"><b>La parall\u00e9lisation<\/b><span style=\"font-weight: 400;\"> permet \u00e0 Apache Flink de traiter les donn\u00e9es en distribuant les t\u00e2ches sur plusieurs n\u0153uds de traitement (des <\/span><b>\u00ab\u00a0slots\u00a0\u00bb<\/b><span style=\"font-weight: 400;\">) simultan\u00e9ment. Cela permet de r\u00e9duire consid\u00e9rablement le temps de traitement des donn\u00e9es en exploitant pleinement les ressources de l\u2019environnement distribu\u00e9. Les t\u00e2ches peuvent \u00eatre divis\u00e9es en plusieurs sous-t\u00e2ches qui sont ensuite ex\u00e9cut\u00e9es en parall\u00e8le, augmentant ainsi l\u2019efficacit\u00e9 de traitement.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>La r\u00e9partition<\/b><span style=\"font-weight: 400;\"> des donn\u00e9es consiste \u00e0 diviser les donn\u00e9es en petits groupes, appel\u00e9s <\/span><b>\u00ab\u00a0partitions\u00a0\u00bb<\/b><span style=\"font-weight: 400;\">, qui sont ensuite distribu\u00e9es sur plusieurs n\u0153uds de traitement. Les partitions peuvent \u00eatre r\u00e9parties en fonction de cl\u00e9s, valeurs, ou tout autre crit\u00e8re de r\u00e9partition afin de permettre de les traiter plus efficacement, \u00e9vitant par cons\u00e9quent les goulots d\u2019\u00e9tranglement. Cette fonctionnalit\u00e9 est un aspect fondamental de la parall\u00e9lisation dans Apache Flink.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Apache Flink est con\u00e7u pour \u00eatre <\/span><b>tol\u00e9rant aux pannes <\/b><span style=\"font-weight: 400;\">gr\u00e2ce \u00e0 un syst\u00e8me de <\/span><b>checkpointing<\/b><span style=\"font-weight: 400;\">, ce qui signifie qu\u2019il peut continuer \u00e0 traiter les donn\u00e9es, m\u00eame en cas de d\u00e9faillance d\u2019un ou de plusieurs n\u0153uds de traitement. Il peut d\u00e9tecter\u00a0 automatiquement les pannes et r\u00e9pliquer les donn\u00e9es en cons\u00e9quence.<\/span><\/li><\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7f32eaa elementor-widget elementor-widget-image\" data-id=\"7f32eaa\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"553\" height=\"200\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image2-2.png\" class=\"attachment-full size-full wp-image-191777\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image2-2.png 553w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image2-2-300x108.png 300w\" sizes=\"(max-width: 553px) 100vw, 553px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-8ba9b6b e-flex e-con-boxed e-con\" data-id=\"8ba9b6b\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-615ca52 elementor-widget elementor-widget-heading\" data-id=\"615ca52\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Comparatif entre Flink et Spark Streaming<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-71772b7 elementor-widget__width-initial elementor-widget elementor-widget-html\" data-id=\"71772b7\" data-element_type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style type=\"text\/css\">\r\n.tg  {border-collapse:collapse;border-spacing:0;}\r\n.tg td{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px;\r\n  overflow:hidden;padding:10px 5px;word-break:normal;}\r\n.tg th{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px;\r\n  font-weight:normal;overflow:hidden;padding:10px 5px;word-break:normal;}\r\n.tg .tg-hojr{background-color:#343434;border-color:inherit;color:#ffffff;font-family:Arial, Helvetica, sans-serif !important;\r\n  font-weight:bold;text-align:left;vertical-align:middle}\r\n.tg .tg-xbtn{background-color:#343434;border-color:inherit;color:#ffffff;font-weight:bold;text-align:center;vertical-align:top}\r\n.tg .tg-84g7{background-color:#343434;border-color:inherit;color:#ffffff;font-family:Arial, Helvetica, sans-serif !important;\r\n  font-weight:bold;text-align:left;vertical-align:top}\r\n.tg .tg-3xi5{background-color:#ffffff;border-color:inherit;text-align:center;vertical-align:top}\r\n.tg .tg-jbyd{background-color:#ffffff;border-color:#000000;text-align:center;vertical-align:top}\r\n<\/style>\r\n<table class=\"tg\">\r\n<thead>\r\n  <tr>\r\n    <th class=\"tg-84g7\">Caract\u00e9ristique<\/th>\r\n    <th class=\"tg-xbtn\">Apache Flink<\/th>\r\n    <th class=\"tg-xbtn\">Spark Streaming<\/th>\r\n  <\/tr>\r\n<\/thead>\r\n<tbody>\r\n  <tr>\r\n    <td class=\"tg-84g7\">Traitement temps r\u00e9el<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<\/td>\r\n  <\/tr>\r\n  <tr>\r\n    <td class=\"tg-hojr\">Traitement de flux<\/td>\r\n    <td class=\"tg-jbyd\">\u2705<br>Traitement de flux infinis avec une latence tr\u00e8s faible<\/td>\r\n    <td class=\"tg-3xi5\">\u2705 <br>Limit\u00e9 par la taille des lots et les temps de latence associ\u00e9s<\/td>\r\n  <\/tr>\r\n  <tr>\r\n    <td class=\"tg-84g7\">Traitement de batchs<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<\/td>\r\n    <td class=\"tg-3xi5\">\u274c<\/td>\r\n  <\/tr>\r\n  <tr>\r\n    <td class=\"tg-hojr\">Mod\u00e8le<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>Bas\u00e9 sur des transformations d'op\u00e9rateurs<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>Bas\u00e9 sur des transformations RDD<\/td>\r\n  <\/tr>\r\n  <tr>\r\n    <td class=\"tg-hojr\">Graphes<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>Via l'API Gelly<\/td>\r\n    <td class=\"tg-3xi5\">\u274c<br>Pas de prise en charge<\/td>\r\n  <\/tr>\r\n  <tr>\r\n    <td class=\"tg-hojr\">Langages<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>Python, Java et Scala<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>Python, Java et Scala<\/td>\r\n  <\/tr>\r\n  <tr>\r\n    <td class=\"tg-hojr\">Tol\u00e9rance de pannes<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>G\u00e8re les pannes de noeuds de traitement<\/td>\r\n    <td class=\"tg-3xi5\">\u2705<br>Tol\u00e9rant aux pannes, mais moins efficacement<\/td>\r\n  <\/tr>\r\n<\/tbody>\r\n<\/table>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-f186c53 e-flex e-con-boxed e-con\" data-id=\"f186c53\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-ca719d2 elementor-widget elementor-widget-heading\" data-id=\"ca719d2\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Quelles sont les limites de Flink ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0e42fee elementor-widget elementor-widget-text-editor\" data-id=\"0e42fee\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Comme nous venons de le voir, Flink est un outil tr\u00e8s puissant pour le traitement de donn\u00e9es. Il a cependant quelques limitations qui sont n\u00e9cessaires de prendre en consid\u00e9ration :\u00a0<\/span><\/p><ul><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Besoins en ressources<\/b><span style=\"font-weight: 400;\"> : Il n\u00e9cessite beaucoup de ressource pour s\u2019ex\u00e9cuter efficacement, notamment pour le stockage de donn\u00e9es et la puissance de traitement<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Complexe \u00e0 mettre en \u0153uvre<\/b><span style=\"font-weight: 400;\"> : Il peut \u00eatre difficile \u00e0 mettre en \u0153uvre pour les d\u00e9veloppeurs ayant peu d\u2019exp\u00e9rience en traitement de donn\u00e9es en temps r\u00e9el. La courbe d\u2019apprentissage peut s&rsquo;av\u00e9rer tr\u00e8s \u00e9lev\u00e9e.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Documentation <\/b><span style=\"font-weight: 400;\">: Sa documentation est compl\u00e8te, mais difficile \u00e0 appr\u00e9hender\u00a0 selon l\u2019exp\u00e9rience du lecteur<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Ecosyst\u00e8me <\/b><span style=\"font-weight: 400;\">: Son d\u00e9veloppement \u00e0 \u00e9t\u00e9 rapide, mais Flink est relativement nouveau par rapport \u00e0 d\u2019autres outils du march\u00e9. Par cons\u00e9quent, il peut manquer de maturit\u00e9 sur certains sujet, ou manquer de fonctionnalit\u00e9s<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><b>Mont\u00e9es de version<\/b><span style=\"font-weight: 400;\"> : Le code d\u00e9velopp\u00e9 sur des versions ant\u00e9rieures ne sont pas forc\u00e9ment compatible apr\u00e8s une mise \u00e0 jour des API\u2019s, ce qui peut s\u2019av\u00e9rer chronophage pour rendre \u00e0 nouveau son code compatible<\/span><\/li><\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-414b0bf e-flex e-con-boxed e-con\" data-id=\"414b0bf\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-f4421ca elementor-widget elementor-widget-heading\" data-id=\"f4421ca\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Conclusion<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f796afd elementor-widget elementor-widget-text-editor\" data-id=\"f796afd\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Pour conclure, Apache Flink est un framework de traitement de donn\u00e9es extr\u00eamement <strong>puissant et polyvalent<\/strong>. Il offre des fonctionnalit\u00e9s avanc\u00e9es pour le <strong>traitement de flux et de batch<\/strong>, et permet aux utilisateurs de <strong>r\u00e9aliser des t\u00e2ches de traitement de donn\u00e9es en temps r\u00e9el<\/strong> de mani\u00e8re efficace et \u00e9volutive.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Les API de programmation de Flink sont faciles \u00e0 utiliser, <strong>offrant une grande flexibilit\u00e9 pour les d\u00e9veloppeurs<\/strong>, et sa capacit\u00e9 \u00e0 g\u00e9rer des charges de travail de grande envergure en fait un choix de premier ordre pour les entreprises et les organisations de toutes tailles.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Bien qu&rsquo;il ait certaines limites, Flink continue d&rsquo;\u00e9voluer rapidement pour r\u00e9pondre aux besoins changeants des entreprises.\u00a0<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-db5a880 elementor-widget elementor-widget-text-editor\" data-id=\"db5a880\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Maintenant que vous savez tout sur Apache Flink, commencez une formation qui vous permette de<strong> ma\u00eetriser enti\u00e8rement cet outil de traitement de flux<\/strong>. D\u00e9couvrez DataScientest !<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b562631 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"b562631\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formations-data\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">D\u00e9marrer une formation en Data Science<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-83866f7 e-flex e-con-boxed e-con\" data-id=\"83866f7\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-0c44ee3 elementor-widget elementor-widget-html\" data-id=\"0c44ee3\" data-element_type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<script type=\"application\/ld+json\">\r\n{\r\n  \"@context\": \"https:\/\/schema.org\",\r\n  \"@type\": \"FAQPage\",\r\n  \"mainEntity\": [{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Comment est structur\u00e9 l'architecture de Flink ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Flink est un moteur de traitement distribu\u00e9 pour des calculs avec \u00e9tat sur des flux de donn\u00e9es born\u00e9s ou illimit\u00e9s. Il a \u00e9t\u00e9 con\u00e7u pour fonctionner dans tous les environnements de cluster courants, et pour \u00eatre en mesure d\u2019effectuer des calculs \u00e0 une vitesse tr\u00e8s \u00e9lev\u00e9e, et \u00e0 n\u2019importe quelle \u00e9chelle.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Quelles sont les applications li\u00e9es \u00e0 Flink gr\u00e2ce aux API\u2019s ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Apache Flink propose un ensemble d\u2019API tr\u00e8s riches permettant de faire des transformations \u00e0 la fois sur des donn\u00e9es par lot (batch) ou en temps-r\u00e9el (streaming). Ces transformations sont effectu\u00e9es sur les donn\u00e9es distribu\u00e9es et permettent aux d\u00e9veloppeurs de construire une application qui r\u00e9pondra \u00e0 leur besoin.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Quelles sont les limites de Flink ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Besoins en ressources : Il n\u00e9cessite beaucoup de ressource pour s\u2019ex\u00e9cuter efficacement, notamment pour le stockage de donn\u00e9es et la puissance de traitement\"\r\n    }\r\n  }]\r\n}\r\n<\/script>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Les origines de Apache Flink Initialement d\u00e9velopp\u00e9 \u00e0 l\u2019universit\u00e9 technique de Berlin, ses premi\u00e8res versions ont \u00e9t\u00e9 publi\u00e9es en 2011 et avaient pour but de r\u00e9pondre \u00e0 des probl\u00e9matiques complexes dans le traitement de donn\u00e9es dans un environnement distribu\u00e9 en temps r\u00e9el. Flink est devenu au fil des ann\u00e9es une r\u00e9f\u00e9rence pour un grand nombre [&hellip;]<\/p>\n","protected":false},"author":78,"featured_media":191771,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[2497],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.7 (Yoast SEO v21.7) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Apache Flink : D\u00e9finition, structure et limites du framework<\/title>\n<meta name=\"description\" content=\"D\u00e9couvrez tout ce qu&#039;il y a \u00e0 savoir \u00e0 propos d&#039;Apache Flink, un moteur de flux de donn\u00e9es open-source d\u00e9velopp\u00e9 par Apache Software Foundation.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/datascientest.com\/apache-flink-tout-savoir\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Apache Flink : D\u00e9finition, structure et limites du framework\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/apache-flink-tout-savoir\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2023-03-08T18:39:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-09T13:53:53+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/apache-flink-datascientest.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"600\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"J\u00e9r\u00e9my Robert\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"J\u00e9r\u00e9my Robert\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/apache-flink-tout-savoir\",\"url\":\"https:\/\/datascientest.com\/apache-flink-tout-savoir\",\"name\":\"Apache Flink : D\u00e9finition, structure et limites du framework\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2023-03-08T18:39:00+00:00\",\"dateModified\":\"2023-11-09T13:53:53+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925\"},\"description\":\"D\u00e9couvrez tout ce qu'il y a \u00e0 savoir \u00e0 propos d'Apache Flink, un moteur de flux de donn\u00e9es open-source d\u00e9velopp\u00e9 par Apache Software Foundation.\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/apache-flink-tout-savoir\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/datascientest.com\/#website\",\"url\":\"https:\/\/datascientest.com\/\",\"name\":\"Formation Data Science | DataScientest.com\",\"description\":\"Leader de la formation en Data Science\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/datascientest.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925\",\"name\":\"J\u00e9r\u00e9my Robert\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg\",\"caption\":\"J\u00e9r\u00e9my Robert\"},\"description\":\"J\u00e9r\u00e9my Robert est actuellement Directeur p\u00e9dagogique B2B chez DataScientest.com. Avant de rejoindre DataScientest, il a travaill\u00e9 en tant que Consultant Business Intelligence chez DEAPDATA, o\u00f9 il a g\u00e9r\u00e9 des projets dans les domaines de l'assurance, bancaire et industrie. Il a \u00e9galement \u00e9t\u00e9 Co-fondateur et pr\u00e9sident de Tripzer. En tant que consultant ind\u00e9pendant, J\u00e9r\u00e9my a propos\u00e9 des formations sur MicroStrategy pour divers niveaux professionnels. Il est dipl\u00f4m\u00e9 de l'Universit\u00e9 Paris Cit\u00e9 avec un Master MIAGE, combinant comp\u00e9tences en informatique et gestion.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/jeremyrobert1\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/jeremyrobert\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Apache Flink : D\u00e9finition, structure et limites du framework","description":"D\u00e9couvrez tout ce qu'il y a \u00e0 savoir \u00e0 propos d'Apache Flink, un moteur de flux de donn\u00e9es open-source d\u00e9velopp\u00e9 par Apache Software Foundation.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/datascientest.com\/apache-flink-tout-savoir","og_locale":"fr_FR","og_type":"article","og_title":"Apache Flink : D\u00e9finition, structure et limites du framework","og_url":"https:\/\/datascientest.com\/apache-flink-tout-savoir","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2023-03-08T18:39:00+00:00","article_modified_time":"2023-11-09T13:53:53+00:00","og_image":[{"width":1200,"height":600,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/apache-flink-datascientest.png","type":"image\/png"}],"author":"J\u00e9r\u00e9my Robert","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"J\u00e9r\u00e9my Robert","Dur\u00e9e de lecture estim\u00e9e":"8 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/apache-flink-tout-savoir","url":"https:\/\/datascientest.com\/apache-flink-tout-savoir","name":"Apache Flink : D\u00e9finition, structure et limites du framework","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2023-03-08T18:39:00+00:00","dateModified":"2023-11-09T13:53:53+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925"},"description":"D\u00e9couvrez tout ce qu'il y a \u00e0 savoir \u00e0 propos d'Apache Flink, un moteur de flux de donn\u00e9es open-source d\u00e9velopp\u00e9 par Apache Software Foundation.","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/apache-flink-tout-savoir"]}]},{"@type":"WebSite","@id":"https:\/\/datascientest.com\/#website","url":"https:\/\/datascientest.com\/","name":"Formation Data Science | DataScientest.com","description":"Leader de la formation en Data Science","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/datascientest.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925","name":"J\u00e9r\u00e9my Robert","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg","caption":"J\u00e9r\u00e9my Robert"},"description":"J\u00e9r\u00e9my Robert est actuellement Directeur p\u00e9dagogique B2B chez DataScientest.com. Avant de rejoindre DataScientest, il a travaill\u00e9 en tant que Consultant Business Intelligence chez DEAPDATA, o\u00f9 il a g\u00e9r\u00e9 des projets dans les domaines de l'assurance, bancaire et industrie. Il a \u00e9galement \u00e9t\u00e9 Co-fondateur et pr\u00e9sident de Tripzer. En tant que consultant ind\u00e9pendant, J\u00e9r\u00e9my a propos\u00e9 des formations sur MicroStrategy pour divers niveaux professionnels. Il est dipl\u00f4m\u00e9 de l'Universit\u00e9 Paris Cit\u00e9 avec un Master MIAGE, combinant comp\u00e9tences en informatique et gestion.","sameAs":["https:\/\/www.linkedin.com\/in\/jeremyrobert1\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/jeremyrobert"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/191769"}],"collection":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/users\/78"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=191769"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/191769\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/191771"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=191769"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=191769"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=191769"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}