[
https://issues.apache.org/jira/browse/LUCENE-3404?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13092306#comment-13092306
]
Michael McCandless commented on LUCENE-3404:
--------------------------------------------
OK I found the problem -- it's because we are incorrectly passing true instead
of false, as the first arg to IOUtils.closeSafely in IW.copySegmentIntoCFS.
This then suppresses an exception (that corrupted the .cfe file being written),
thus making the rest of the code think everything is fine...
I looked at all other places we call this method and there were a few other
cases.
I think this is trappy -- instead of a boolean, I think we should rename
IOUtils.closeSafely into 2 methods: closeSafelyDuringExc, and closeSafely.
> testIWondiskfull checkindex failure
> -----------------------------------
>
> Key: LUCENE-3404
> URL: https://issues.apache.org/jira/browse/LUCENE-3404
> Project: Lucene - Java
> Issue Type: Bug
> Reporter: Robert Muir
>
> looks like charlie cron created a corrupt index on disk full.. can't
> reproduce with the seed on this machine, i can try on that VM with the same
> environment and see if i have better luck.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
