[
https://issues.apache.org/jira/browse/FLUME-865?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13207488#comment-13207488
]
Juhani Connolly commented on FLUME-865:
---------------------------------------
Ok, I understand your concerns... I perhaps we can break off the processor
part into a separate patch. Honestly I think both approaches (mine and yours)
are seriously hampered by having to maintain backwards compatibility with
existing configurations, which is not necessarily a requirement in this alpha
stage, perhaps we should give consideration to approaches that would alter the
current configuration but result in a more elegant, easy to understand and
implement configuration?
Before that however I would like to know how you suggest we handle sink
specific configurations, namely priority(in a failover processor) and weight(in
a load balancing processor), and possible others for more advanced processors.
Additionally, how do you suggest we handle the groups? Adding:
- List<ComponentConfiguration> AgentConfiguration.getSinkGroups() which would
be called and processed in PropertiesFileConfiguration->loadSinks?
Incidentally, I am somewhat confused by why we have the loadSinks logic in
PropertiesFileConfiguration... Should this not be in a higher layer of
abstraction, non-specific to file based configuration?
After processing all sink groups and assigning sinks to them, then we would
assign any leftover sinks to default processors?
I will take some more time to look at the configuration stuff in detail, and
hope to hear your thoughts on these issues.
> Implement failover sink
> ------------------------
>
> Key: FLUME-865
> URL: https://issues.apache.org/jira/browse/FLUME-865
> Project: Flume
> Issue Type: New Feature
> Components: Sinks+Sources
> Affects Versions: NG alpha 2
> Reporter: Jarek Jarcec Cecho
> Assignee: Juhani Connolly
> Fix For: v1.1.0
>
> Attachments: FLUME-865.patch
>
>
> It would be nice if the flume-ng would have ability to failover to different
> sink in case that the active one is not responding (e.g. before failing the
> transaction).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
