jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefan Guggisberg <stefan.guggisb...@gmail.com>
Subject Re: How to remove corrupted node from repository?
Date Fri, 07 Aug 2009 16:22:43 GMT
On Fri, Aug 7, 2009 at 4:54 PM, Jan Grathwohl<jan.grathwohl@kontrast.de> wrote:
> Hi Stefan,
>> what's the exact output of the consistencyCheck/Fix? dangling child node
>> entries
>> should be fixable.
> The output was :
> 2009-08-06 12:01:39,350 ERROR
> [org.apache.jackrabbit.core.persistence.bundle.BundleDbPersistenceManager]
> NodeState cd0adbec-b2
> 7d-44b3-900f-d720f705c2e6 references inexistent parent id
> c511b2ee-8183-43ea-ae96-fabb27badfba

hmm, that's not the output of the consistencyCheck/Fix code.
seems like the consistencyCheck code didn't run.

did you specify the consistencyCheck/Fix option attribute in the
repository.xml only? you need to specify the option in the
workspace.xml file.

> If I understand the message right, this is a different node that references
> the same inexistent parent as the node causing the problem? At least the
> first UUID ist not mentioned in the exception message...
>> do you know the path of the corrupted node? are you able to navigate to
>> the node?
> No, but getting the node by UUID still works...
>>> Is there anything that I could do to solve this or remove the node, maybe
>>> through Jackrabbit API or even direct manipulation of the database
>>> content?
>>> Any ideas are highly appreciated.
>> if everything fails, you could try to fix that inconsistency
>> programmatically,
>> e.g. by directly reading the troublesome NodeState instance through the
>> pm, fix it and write it back. but you'll have to get your hands dirty...
> That's what I did now. Wrote a small program that accesses the workspace
> persistence manager through the JCA handle and RepositoryImpl (partially by
> using Java reflection to call protected methods and fields), and then
> deleted the broken NodeState from the repository. So far that doesn't seem
> to have caused any further problems.

excellent :)


> Many thanks for the hint!
> Regards,
> Jan

View raw message