lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Uwe Schindler (JIRA)" <j...@apache.org>
Subject [jira] [Created] (SOLR-6518) CachingDirectoryFactory subclasses must init directory with NoLockFactory, because the real lock factory gets set later via injectLockFactory
Date Mon, 15 Sep 2014 00:45:33 GMT
Uwe Schindler created SOLR-6518:
-----------------------------------

             Summary: CachingDirectoryFactory subclasses must init directory with NoLockFactory,
because the real lock factory gets set later via injectLockFactory
                 Key: SOLR-6518
                 URL: https://issues.apache.org/jira/browse/SOLR-6518
             Project: Solr
          Issue Type: Bug
    Affects Versions: 5.0
            Reporter: Uwe Schindler
            Assignee: Uwe Schindler
             Fix For: 5.0


The whole setup should really be changed in Solr trunk, especially as we want to make the
lockfactory non-mutable (set via directory ctor).

As workaround to prevent SimpleFSLockFactory from creating the lock directory initially, we
should pass NoLockFactory in the create method. This is possible because injectLockFactory
will set the "configured one" afterwards.

This change is needed in trunk, because the LockFactory creates the directory in its ctor,
which leads to problems with NIO2.

In addition, the SlrCore.newSolrDataDir does not cleanup and close the directory correctly
if an error occurs.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message