[
https://issues.apache.org/jira/browse/CURATOR-57?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13778553#comment-13778553
]
Benjamin Jaton commented on CURATOR-57:
---------------------------------------
Well I think there is/has been some confusion about internalRebuildNode().
It is up to the PathChildrenCache implementation to retrieve the Stat object
independently or not. We could imagine another implementation that would have
preferred to iterate over the result of the getChildren() watcher, wait for all
the corresponding Stat objects to be gathered before making an informed
decision on which events should be dispatched first.
I agree that this could be achieved in another recipe, based on
PathChildrenCache but that would be single threaded (makes no sense to dispatch
the events in parallel otherwise we could lose the ordering).
> Event order lost with PathChildrenCache
> ---------------------------------------
>
> Key: CURATOR-57
> URL: https://issues.apache.org/jira/browse/CURATOR-57
> Project: Apache Curator
> Issue Type: Bug
> Components: Recipes
> Affects Versions: 2.0.1-incubating
> Environment: Ubuntu 10.10
> Reporter: Benjamin Jaton
> Fix For: awaiting-response
>
> Attachments: Test.java
>
>
> I am attaching a simple test to demonstrate the problem:
> Single threaded 'for' loop that creates children 0 to N, the events are not
> received in order.
--
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
