[
https://issues.apache.org/jira/browse/FLUME-718?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13078867#comment-13078867
]
E. Sammer commented on FLUME-718:
---------------------------------
I definitely don't like the idea of having another layer of "defaults" hiding
in the code. Hadoop has these issues as well. I believe the correct thing to do
here is to drop a copy of flume-conf.xml in src/main/resources so it winds up
in the jar and guarantee that all params are set here (so the null condition
never happens in FlumeConfiguration). As it is today, there's three places a
param can be set (or not set), FlumeConfiguration not being readily
accessible[1] to an end user, and that's just begging for trouble.
Let's not do this.
[1] By which I mean the source should not be the docs for a non-programmer ops
type of person.
> FlumeConfiguration should have default webapp paths.
> ----------------------------------------------------
>
> Key: FLUME-718
> URL: https://issues.apache.org/jira/browse/FLUME-718
> Project: Flume
> Issue Type: Bug
> Affects Versions: v0.9.4
> Reporter: Jonathan Hsieh
> Assignee: Jonathan Hsieh
> Fix For: v0.9.5
>
> Attachments:
> 0001-FLUME-71x-FlumeConfiguration-should-have-default-web.patch
>
>
> The FlumeConfiguration value should have a reasonable default in the case
> that configuration files do not have any value. As an example, 0.9.3 config
> files do not have the flume.master.webapp.root values and if used after an
> upgrade to 0.9.4, flume's status webpages will fail.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
