[
https://issues.apache.org/jira/browse/SOLR-3699?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13445143#comment-13445143
]
Hoss Man commented on SOLR-3699:
--------------------------------
bq. Yeah, brain murmur - for some reason I was thinking DirectoryFactory was
used globally across cores.
I think it's only shared between reloaded versions of the same SolrCore (ie:
two differnet SolrCore's can use completley diff DirectoryFactory impls) but
even then, it's only shared if the SolrCore was constructed using an existing
UpdateHandler (ie: reload) if the SolrCore is the "first" SolrCore, then it
creates the DirectoryFactory, thus if the UpdateHandler doesn't init properly,
that SolrCore needs to close the DirectoryFactory.
Ideally, a whole metric-shit-load of the SolrCore initialization would be
reworked, so thta every type of object (like DirectoryFactory) was only ever
inited by one other object (like UpdateHandler) regardless of wether it's the
first create or reload state, ... but i wasn't really comfortable making that
heavy of a change.
> SolrIndexWriter constructor leaks Directory if Exception creating
> IndexWriterConfig
> -----------------------------------------------------------------------------------
>
> Key: SOLR-3699
> URL: https://issues.apache.org/jira/browse/SOLR-3699
> Project: Solr
> Issue Type: Bug
> Reporter: Robert Muir
> Assignee: Mark Miller
> Fix For: 4.0
>
> Attachments: SOLR-3699.patch, SOLR-3699.patch, SOLR-3699.patch,
> SOLR-3699.patch
>
>
> in LUCENE-4278 i had to add a hack to force SimpleFSDir for
> CoreContainerCoreInitFailuresTest, because it doesnt close its Directory on
> certain errors.
> This might indicate a problem that leaks happen if certain errors happen
> (e.g. not handled in finally)
--
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
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
