You are viewing a plain text version of this content. The canonical link for it is here.

Posted to common-dev@hadoop.apache.org by "Robert Chansler (JIRA)" <ji...@apache.org> on 2008/08/22 22:24:44 UTC

[jira] Created: (HADOOP-4004) NameNode should not serve up a bad edits log

NameNode should not serve up a bad edits log
--------------------------------------------

                 Key: HADOOP-4004
                 URL: https://issues.apache.org/jira/browse/HADOOP-4004
             Project: Hadoop Core
          Issue Type: Bug
    Affects Versions: 0.17.1
            Reporter: Robert Chansler


A NameNode disk failure (apparently) resulted in the NameNode serving a bad edits log to the Secondary NameNode. The SNN observed the problem (good!), but had no alternative but to ask again for the log, and again get the same bad replica.

1. The NN could/should have observed the same fault as the SNN.
2. If a replica is known to be bad, the NN should serve a different replica, if available.
3. The SNN should have a way to report replica failure to the NN.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

[jira] Updated: (HADOOP-4004) NameNode should not serve up a bad edits log

Posted by "Robert Chansler (JIRA)" <ji...@apache.org>.

     [ https://issues.apache.org/jira/browse/HADOOP-4004?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Robert Chansler updated HADOOP-4004:
------------------------------------

    Component/s: dfs

> NameNode should not serve up a bad edits log
> --------------------------------------------
>
>                 Key: HADOOP-4004
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4004
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: dfs
>    Affects Versions: 0.17.1
>            Reporter: Robert Chansler
>
> A NameNode disk failure (apparently) resulted in the NameNode serving a bad edits log to the Secondary NameNode. The SNN observed the problem (good!), but had no alternative but to ask again for the log, and again get the same bad replica.
> 1. The NN could/should have observed the same fault as the SNN.
> 2. If a replica is known to be bad, the NN should serve a different replica, if available.
> 3. The SNN should have a way to report replica failure to the NN.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.