Mailing-List: contact dev-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@jackrabbit.apache.org
Date: Mon, 28 Jan 2013 08:31:13 +0000 (UTC)
From: "Unico Hommes (JIRA)" <jira@apache.org>
To: dev@jackrabbit.apache.org
Message-ID: <JIRA.12629449.1359316973292.200003.1359361873065@arcas>
In-Reply-To: <JIRA.12629449.1359316973292@arcas>
References: <JIRA.12629449.1359316973292@arcas>
Subject: [jira] [Updated] (JCR-3501) When cancelling an update modcount of
 modified states must be reset
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


     [ https://issues.apache.org/jira/browse/JCR-3501?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Unico Hommes updated JCR-3501:
------------------------------

    Fix Version/s: 2.7
                   2.6
                   2.4.4
                   2.2.14
    
> When cancelling an update modcount of modified states must be reset
> -------------------------------------------------------------------
>
>                 Key: JCR-3501
>                 URL: https://issues.apache.org/jira/browse/JCR-3501
>             Project: Jackrabbit Content Repository
>          Issue Type: Bug
>            Reporter: Unico Hommes
>            Assignee: Unico Hommes
>            Priority: Critical
>             Fix For: 2.2.14, 2.4.4, 2.6, 2.7
>
>
> The overlayed states are touched when beginning an update, but when an update is cancelled they are never untouched again. A subsequent external change to the same state is not merged correctly because the item will not appear stale.
> The problem causes data inconsistencies in the following example scenario:
> remote.getNode("/test").addNode("a");
> remote.save();
> local.getNode("/test").addNode("b");
> local.save();
> local.getNode("/test").getNode("a").addNode("c");
> local.save();
> remote.getNode("/test").getNode("a").remove();
> final Node d = local.getNode("/test").addNode("d");
> try {
>     remote.save(); // fails, but cancelling update leaves SISM in bad state
> } catch (Exception e) {
>     remote.refresh(false);
> }
> remote.getNode("/test").addNode("e");
> local.save();
> remote.save(); // succeeds but leaves repository inconsistencies

--
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