Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hdfs-issues@hadoop.apache.org
Date: Fri, 12 Jul 2013 18:35:54 +0000 (UTC)
From: "Kihwal Lee (JIRA)" <jira@apache.org>
To: hdfs-issues@hadoop.apache.org
Message-ID: <JIRA.12502074.1300818558475.41176.1373654154039@arcas>
In-Reply-To: <JIRA.12502074.1300818558475@arcas>
References: <JIRA.12502074.1300818558475@arcas>
Subject: [jira] [Resolved] (HDFS-1775) getContentSummary should hold the
 FSNamesystem readlock
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


     [ https://issues.apache.org/jira/browse/HDFS-1775?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Kihwal Lee resolved HDFS-1775.
------------------------------

    Resolution: Duplicate
    
> getContentSummary should hold the FSNamesystem readlock
> -------------------------------------------------------
>
>                 Key: HDFS-1775
>                 URL: https://issues.apache.org/jira/browse/HDFS-1775
>             Project: Hadoop HDFS
>          Issue Type: Bug
>            Reporter: Dmytro Molkov
>            Assignee: Dmytro Molkov
>            Priority: Minor
>
> Right now the getContentSummary call on the namenode only holds the FSDirectory lock, but not the FSNamesystem lock. What we are seeing because of that is: 
> 1) getContentSummary takes the read lock on FSDirectory 
> 2) the write operation comes and takes a write lock on FSNamesystem and waits for getContentSummary to finish to get a write lock on FSDirectory
> As a result other read operations can't be executed. Since getContentSummary can take a while to execute on large directories, the performance would be improved if we hold a readlock while doing that.

--
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