[jira] [Updated] (ZOOKEEPER-1483) Fix leader election recipe documentation

["Michi Mutsuzaki (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/ZOOKEEPER-1483?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Michi Mutsuzaki updated ZOOKEEPER-1483:
---------------------------------------

    Attachment: ZOOKEEPER-1483-trunk.patch
                ZOOKEEPER-1483-3.4.patch

Created patches for trunk and 3.4 branch. In addition to the recipe error, These patches fix a grammatical error.
                
> Fix leader election recipe documentation
> ----------------------------------------
>
>                 Key: ZOOKEEPER-1483
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-1483
>             Project: ZooKeeper
>          Issue Type: Bug
>          Components: documentation
>    Affects Versions: 3.4.3
>            Reporter: Ankur
>            Assignee: Michi Mutsuzaki
>             Fix For: 3.4.4, 3.5.0
>
>         Attachments: ZOOKEEPER-1483-3.4.patch, ZOOKEEPER-1483-trunk.patch
>
>
> The leader election recipe documentation suggest that to avoid the herd effect a client process volunteering for leadership via child znode [i] under the leader election path [/leader] must only watch the the SMALLEST znode [j] from a different client process such that [j < i]. 
> This will NOT avoid the herd effect as many clients will end up watching the same znode[j] where j is the next-in-sequence number greater than the number of the current leader.
> Specifically in Step 3 of the Election procedure here http://zookeeper.apache.org/doc/trunk/recipes.html#sc_leaderElection
> This "where j is the SMALLEST sequence number" should be changed to this
> "where j is the LARGEST sequence number"

--
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