[
https://issues.apache.org/jira/browse/S4-113?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13555266#comment-13555266
]
Matthieu Morel commented on S4-113:
-----------------------------------
Thanks for pointing this out! In some of the checkpointing tests, we rely on
Zookeeper notifications. Unfortunately, (my bad), these notifications can be
received before file checkpoints are completed (especially with slow disks).
Adding pauses is far from ideal, though apart from intercepting the checkpoint
notification (and thus modifying the classes we are testing), or using java 7's
improved file API, it might be the best solution for the moment. So I'll do
that, but any other suggestion is welcome.
> Recovery tests fail sometimes
> -----------------------------
>
> Key: S4-113
> URL: https://issues.apache.org/jira/browse/S4-113
> Project: Apache S4
> Issue Type: Bug
> Reporter: Daniel Gómez Ferro
> Assignee: Matthieu Morel
>
> The following tests fail sometimes on a somewhat older machine:
> org.apache.s4.core.ft.RecoveryTest
> org.apache.s4.core.ft.FTWordCountTest
--
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
