Mailing-List: contact hdfs-dev-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hdfs-dev@hadoop.apache.org
Date: Sat, 20 Feb 2016 18:06:18 +0000 (UTC)
From: "Arpit Agarwal (JIRA)" <jira@apache.org>
To: hdfs-dev@hadoop.apache.org
Message-ID: <JIRA.12940808.1455991548000.101916.1455991578150@Atlassian.JIRA>
In-Reply-To: <JIRA.12940808.1455991548000@Atlassian.JIRA>
References: <JIRA.12940808.1455991548000@Atlassian.JIRA>
 <JIRA.12940808.1455991548891@arcas>
Subject: [jira] [Created] (HDFS-9839) Reduce verbosity of processReport
 logging
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Arpit Agarwal created HDFS-9839:
-----------------------------------

             Summary: Reduce verbosity of processReport logging
                 Key: HDFS-9839
                 URL: https://issues.apache.org/jira/browse/HDFS-9839
             Project: Hadoop HDFS
          Issue Type: Bug
          Components: namenode
    Affects Versions: 2.8.0
            Reporter: Arpit Agarwal
            Assignee: Arpit Agarwal


{{BlockManager#processReport}} logs one line for each invalidated block at INFO. HDFS-7503 moved this logging outside the NameSystem write lock but we still see the NameNode being slowed down when the number of block invalidations is very large e.g. just after a large amount of data is deleted.

{code}
      for (Block b : invalidatedBlocks) {
        blockLog.info("BLOCK* processReport: {} on node {} size {} does not " +
            "belong to any file", b, node, b.getNumBytes());
      }
{code}

We can change this statement to DEBUG and just log the number of block invalidations at INFO.


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)