jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Thomas Mueller (JIRA)" <j...@apache.org>
Subject [jira] Commented: (JCR-2063) FileDataStore: garbage collection can delete files that are still needed
Date Tue, 16 Feb 2010 13:57:27 GMT

    [ https://issues.apache.org/jira/browse/JCR-2063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12834216#action_12834216
] 

Thomas Mueller commented on JCR-2063:
-------------------------------------

A workaround for implementations where this is not fixed is:

gc.mark();
try {
    // sleep to ensure the last modified time is updated
    // even for file system with a lower time resolution
    Thread.sleep(5000);
} catch (Exception e) {
    // can not ignore, otherwise data that is in use may be deleted
    throw new RepositoryException("Interrupted");
}
gc.mark();



> FileDataStore: garbage collection can delete files that are still needed
> ------------------------------------------------------------------------
>
>                 Key: JCR-2063
>                 URL: https://issues.apache.org/jira/browse/JCR-2063
>             Project: Jackrabbit Content Repository
>          Issue Type: Bug
>          Components: jackrabbit-core
>            Reporter: Thomas Mueller
>            Assignee: Thomas Mueller
>             Fix For: 1.5.5
>
>
> It looks like the FileDataStore garbage collection (both regular scan and persistence
manager scan) can delete files that are still needed.
> Currently it looks like the reason is the last access time resolution of the operating
system. This is 2 seconds for FAT and Mac OS X, NTFS 100 ns, and 1 second for other file systems.
That means file that are scanned at the very beginning are sometimes deleted, because they
have a later last modified time then when the scan was started.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message