Return-Path: X-Original-To: apmail-hadoop-yarn-issues-archive@minotaur.apache.org Delivered-To: apmail-hadoop-yarn-issues-archive@minotaur.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id 0E4A818200 for ; Thu, 17 Mar 2016 13:47:52 +0000 (UTC) Received: (qmail 81929 invoked by uid 500); 17 Mar 2016 13:47:50 -0000 Delivered-To: apmail-hadoop-yarn-issues-archive@hadoop.apache.org Received: (qmail 80948 invoked by uid 500); 17 Mar 2016 13:47:49 -0000 Mailing-List: contact yarn-issues-help@hadoop.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: yarn-issues@hadoop.apache.org Delivered-To: mailing list yarn-issues@hadoop.apache.org Received: (qmail 80485 invoked by uid 99); 17 Mar 2016 13:47:49 -0000 Received: from arcas.apache.org (HELO arcas) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 17 Mar 2016 13:47:49 +0000 Received: from arcas.apache.org (localhost [127.0.0.1]) by arcas (Postfix) with ESMTP id 8709B2C1F6A for ; Thu, 17 Mar 2016 13:47:33 +0000 (UTC) Date: Thu, 17 Mar 2016 13:47:33 +0000 (UTC) From: "Junping Du (JIRA)" To: yarn-issues@hadoop.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Commented] (YARN-4815) ATS 1.5 timelineclinet impl try to create attempt directory for every event call MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394 [ https://issues.apache.org/jira/browse/YARN-4815?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15199541#comment-15199541 ] Junping Du commented on YARN-4815: ---------------------------------- bq. My concern is that we do not need to have separate caches for each use cases if they can be modeled by Guava. I see. There are pros and cons to use third party library like Guava. The good thing is it simply implementation, like: we don't need to handle synchronization in this case. However, the general concern is the consistent (interface and behavior) of APIs is not that promising (even Java API has this problem but we have no other choices), also it is hard to follow bug fix across different versions. I don't have strong preference to use Guava or not. But this shouldn't be the reason to keep patch get in as no consensus in community yet. Any other comments on patch? > ATS 1.5 timelineclinet impl try to create attempt directory for every event call > -------------------------------------------------------------------------------- > > Key: YARN-4815 > URL: https://issues.apache.org/jira/browse/YARN-4815 > Project: Hadoop YARN > Issue Type: Sub-task > Components: timelineserver > Reporter: Xuan Gong > Assignee: Xuan Gong > Attachments: YARN-4815.1.patch > > > ATS 1.5 timelineclinet impl, try to create attempt directory for every event call. Since per attempt only one call to create directory is enough, this is causing perf issue. -- This message was sent by Atlassian JIRA (v6.3.4#6332)