hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HDFS-2079) 1073: create an escape hatch to ignore startup consistency problems
Date Thu, 16 Jun 2011 23:55:47 GMT

     [ https://issues.apache.org/jira/browse/HDFS-2079?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Todd Lipcon updated HDFS-2079:
------------------------------

          Component/s: name-node
          Description: 
In the 1073 branch, the NN is much more careful on startup to make sure it isn't missing any
transactions, etc. For example, if 
one of the log segments is missing, it will refuse to start up with a message like "Expected
next log group would start at txid 502 but starts at txid 516". It would be nice to provide
a last-ditch escape hatch, similar to InnoDB's [force recovery modes|http://dev.mysql.com/doc/refman/5.1/en/forcing-innodb-recovery.html]
to allow the operator to tell the NN to attempt to push past any such problems, with data
loss expected.
    Affects Version/s: Edit log branch (HDFS-1073)
        Fix Version/s: Edit log branch (HDFS-1073)

> 1073: create an escape hatch to ignore startup consistency problems
> -------------------------------------------------------------------
>
>                 Key: HDFS-2079
>                 URL: https://issues.apache.org/jira/browse/HDFS-2079
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: name-node
>    Affects Versions: Edit log branch (HDFS-1073)
>            Reporter: Todd Lipcon
>            Assignee: Todd Lipcon
>             Fix For: Edit log branch (HDFS-1073)
>
>
> In the 1073 branch, the NN is much more careful on startup to make sure it isn't missing
any transactions, etc. For example, if 
> one of the log segments is missing, it will refuse to start up with a message like "Expected
next log group would start at txid 502 but starts at txid 516". It would be nice to provide
a last-ditch escape hatch, similar to InnoDB's [force recovery modes|http://dev.mysql.com/doc/refman/5.1/en/forcing-innodb-recovery.html]
to allow the operator to tell the NN to attempt to push past any such problems, with data
loss expected.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message