Return-Path: X-Original-To: apmail-hadoop-hdfs-issues-archive@minotaur.apache.org Delivered-To: apmail-hadoop-hdfs-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 E250010C55 for ; Tue, 4 Mar 2014 23:31:45 +0000 (UTC) Received: (qmail 61980 invoked by uid 500); 4 Mar 2014 23:31:43 -0000 Delivered-To: apmail-hadoop-hdfs-issues-archive@hadoop.apache.org Received: (qmail 61916 invoked by uid 500); 4 Mar 2014 23:31:43 -0000 Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: hdfs-issues@hadoop.apache.org Delivered-To: mailing list hdfs-issues@hadoop.apache.org Received: (qmail 61905 invoked by uid 99); 4 Mar 2014 23:31:43 -0000 Received: from arcas.apache.org (HELO arcas.apache.org) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 04 Mar 2014 23:31:43 +0000 Date: Tue, 4 Mar 2014 23:31:43 +0000 (UTC) From: "Aaron T. Myers (JIRA)" To: hdfs-issues@hadoop.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Commented] (HDFS-6050) NFS OpenFileCtx does not handle exceptions correctly 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/HDFS-6050?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13920209#comment-13920209 ] Aaron T. Myers commented on HDFS-6050: -------------------------------------- bq. For the config variable names, some start with dfs.nfs and some with dfs.nfs3. The reason is that, we may add nfsv4 (borrowing more of Brock's code ) into the code base eventually. With that in consideration, dfs.nfs.xxx is the configuration shared by both v3 and v4, while dfs.nfs3.xxx and dfs.nfs4.xxx are version specific settings. Also, hadoop.nfs.xxx could be the config variable names in hadoop-common project. Please let me know if this makes sense. I understand in the abstract where you're coming from, but here are a few reasons I think we should go with a more consistent naming scheme. # Even though this might make sense to HDFS developers, it may not in fact be helpful for users. Consistency is probably more important here. I've already had internal testers/users here goof up their setups by having configs that weren't being read because they incorrectly omitted or included "3" at the end of "nfs" in their setting names. # Since all of these settings are by definition server side, there's no strict _need_ for these configs to be in separate namespaces. An operator could entirely reasonably just run their NFS3 gateway against a different config directory than their NFS4 gateway, even if they were on the same box. # It's not at all clear to me why some things were chosen to have "nfs3" and others just "nfs". For example, why was "dfs.nfs.exports.allowed.hosts" chosen to be shared between both 3 and 4, while "dfs.nfs3.export.point" was chosen to be specific to just NFS3? Given all this, I think we should just switch to consistently using the "dfs.nfs" prefix. Most or all of these will in fact be usable by an NFS4 implementation as well if/when that's implemented, and until then including "3" in the config names adds no value. > NFS OpenFileCtx does not handle exceptions correctly > ---------------------------------------------------- > > Key: HDFS-6050 > URL: https://issues.apache.org/jira/browse/HDFS-6050 > Project: Hadoop HDFS > Issue Type: Bug > Components: nfs > Reporter: Brock Noland > Assignee: Brandon Li > > I noticed this file does not log exceptions appropriately in multiple locations. > Not logging the stack of Throwable: > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L364 > Printing exceptions to stderr: > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L1160 > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L1149 > Not logging the stack trace: > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L1062 > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L966 > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L961 > https://github.com/apache/hadoop-common/blob/f567f09091368fe800f3f70605da38a69c953fe3/hadoop-hdfs-project/hadoop-hdfs-nfs/src/main/java/org/apache/hadoop/hdfs/nfs/nfs3/OpenFileCtx.java#L680 -- This message was sent by Atlassian JIRA (v6.2#6252)