Return-Path: X-Original-To: apmail-hadoop-common-issues-archive@minotaur.apache.org Delivered-To: apmail-hadoop-common-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 73B7CD11A for ; Tue, 13 Nov 2012 09:17:15 +0000 (UTC) Received: (qmail 40742 invoked by uid 500); 13 Nov 2012 09:17:15 -0000 Delivered-To: apmail-hadoop-common-issues-archive@hadoop.apache.org Received: (qmail 40607 invoked by uid 500); 13 Nov 2012 09:17:15 -0000 Mailing-List: contact common-issues-help@hadoop.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: common-issues@hadoop.apache.org Delivered-To: mailing list common-issues@hadoop.apache.org Received: (qmail 40505 invoked by uid 99); 13 Nov 2012 09:17:13 -0000 Received: from arcas.apache.org (HELO arcas.apache.org) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 13 Nov 2012 09:17:13 +0000 Date: Tue, 13 Nov 2012 09:17:13 +0000 (UTC) From: "Jing Zhao (JIRA)" To: common-issues@hadoop.apache.org Message-ID: <1767872863.106546.1352798233690.JavaMail.jiratomcat@arcas> In-Reply-To: <79078274.85721.1347821047421.JavaMail.jiratomcat@arcas> Subject: [jira] [Commented] (HADOOP-8820) Backport HADOOP-8469 and HADOOP-8470: add "NodeGroup" layer in new NetworkTopology (also known as NetworkTopologyWithNodeGroup) 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/HADOOP-8820?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13496055#comment-13496055 ] Jing Zhao commented on HADOOP-8820: ----------------------------------- bq. The 1st is not a issue as children is already checked before. I'm not sure if I understand the code correctly here. So the isRack() function first checks the children of the current node ("if (getChildren().isEmpty())"), but does not check the children of the first child of the current node. Thus is it possible that "((InnerNode) firstChild).children" is empty (in which case the following get(0) may cause an IndexOutofBoundsException)? Or somewhere outside the function it has been checked? By the way, in the latest patch in HADOOP-8817 (HADOOP-8817-v3.patch), several method definitions seem to be repeated between BlockPlacementPolicy.java and BlockPlacementPolicyDefault.java (e.g., adjustSetsWithChosenReplica(), splitNodesWithLocalityGroup(), getLocalityGroupForSplit() and getRack()). So do we need to recheck the patch to verify? > Backport HADOOP-8469 and HADOOP-8470: add "NodeGroup" layer in new NetworkTopology (also known as NetworkTopologyWithNodeGroup) > ------------------------------------------------------------------------------------------------------------------------------- > > Key: HADOOP-8820 > URL: https://issues.apache.org/jira/browse/HADOOP-8820 > Project: Hadoop Common > Issue Type: New Feature > Components: net > Affects Versions: 1.0.0 > Reporter: Junping Du > Assignee: Junping Du > Attachments: HADOOP-8820.b1.002.patch, HADOOP-8820.b1.003.patch, HADOOP-8820.patch > > > This patch backport HADOOP-8469 and HADOOP-8470 to branch-1 and includes: > 1. Make NetworkTopology class pluggable for extension. > 2. Implement a 4-layer NetworkTopology class (named as NetworkTopologyWithNodeGroup) to use in virtualized environment (or other situation with additional layer between host and rack). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira