You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@chukwa.apache.org by "Jerome Boulon (JIRA)" <ji...@apache.org> on 2009/12/03 22:06:20 UTC

[jira] Updated: (CHUKWA-87) Rewrite all rolling classes

     [ https://issues.apache.org/jira/browse/CHUKWA-87?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jerome Boulon updated CHUKWA-87:
--------------------------------

    Component/s:     (was: Data Processors)
                 MR Data Processors

> Rewrite all rolling classes
> ---------------------------
>
>                 Key: CHUKWA-87
>                 URL: https://issues.apache.org/jira/browse/CHUKWA-87
>             Project: Hadoop Chukwa
>          Issue Type: Improvement
>          Components: MR Data Processors
>            Reporter: Jerome Boulon
>            Assignee: Jerome Boulon
>
> There's a dependency between Hourly and Daily rolling. 
> This should be simplified to only be a generic compacter.  
> The compacter should run whenever there's more than xx spills files in the stream directory and this could happen any time during the day.
> This may implies flattering the stream hierarchy but the logic for that is only on the 2 rolling classes and the MoveToRepo class.
> The compacter should be able to get the list of files, compact them, put the new file in place and removed the one that have been part of the compacter M/R job.
> The compacter will be a daemon process that will always run in background. 

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