展开
!function(){"use strict";var e=function(e,a){function t(e,a){var t=e.match(new RegExp(a "\s*=\s*["']?([^"'\s>] )["']?"));return t&&t[1]}e=a?window.__emojiFormat(function(e){var a;e=e.split(/(<as w[^>]*>)(.*?)(</a>)/);for(var i=0;i<e.length;i )if(i%4==1){a=!1;var n=t(e[i],"href");/^https?://mp.weixin.qq.com/.test(n)&&(a=!0);var s='<a target="_blank" data-href="' (a?n:"javascript:void 0") '"',r=t(e[i],"data-miniprogram-appid");if(r)s =' class="weapp_text_link" data-miniprogram-type="text" data-miniprogram-appid="' r '" data-miniprogram-path="' (t(e[i],"data-miniprogram-path")||"") '" data-miniprogram-nickname="' (t(e[i],"data-miniprogram-nickname")||"") '" data-miniprogram-servicetype="' (t(e[i],"data-miniprogram-servicetype")||"0") '"',a=!0;else if(a){s =' class="js_video_desc_link"';var o=t(e[i],"data-itemshowtype");o&&(s =' data-itemshowtype="' o '"')}e[i]=a?s ">":""}else i%4==3?!a&&(e[i]=""):e[i]=e[i].replace(/<.*?>/g,"");return e.join("")}(e).replace(/r/g,"").replace(/n/g,"<br>")):e.replace(/r/g,"").replace(/n/g,"<br>").replace(/s/g," ");var i=document.getElementById("js_common_share_desc"),n=document.getElementById("js_common_share_desc1");i&&n&&(i.innerHTML=e,n.innerHTML=e,i.style.maxHeight="",i.classList?!i.classList.contains("default")&&i.classList.add("default"):i.className.match(/(s|^)default(s|$)/)||(i.className =" default"),setTimeout((function(){var e=document.getElementById("js_folder_text_switch");n.offsetHeight>i.offsetHeight 1?e.style.display="block":e.style.display="none"}),300))};if(!window.__second_open__){var a="Apache Impala是用于处理存储在Hadoop中海量数据的MPP架构的SQL查询引擎，用 C 和 Java 编写而成，与其它SQL-on-Hadoop如Hive/Spark相比，它提供更高的性能和更低的延迟。这一次发布的大版本更新Impala 4.0增加了很多新特性和重大改进如所有operator多线程加速，Hive full-ACID支持，Apache Iceberg支持，TPC-DS 99个查询的支持等。本次分享将回顾Impala的发展历程，并简要介绍Impala的基本实现，在此基础上介绍Impala 4.0的主要更新，以及未来的技术路线图。";e(a,!0),window.__setDesc=e}}();
