jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dominique Pfister" <dominique.pfis...@day.com>
Subject Re: Clustering documentation
Date Mon, 07 May 2007 15:30:07 GMT
Hi Andrea,

On 5/7/07, Andrea K. <akregar@alteanet.it> wrote:
> thanks a lot for the answers and the great wiki page. I hope to give some
> support on it soon.

That would be great!

> Just a question about JCA connector: I assume I have to deploy it on every
> node, is it right?

That is right. Remember, to assign unique cluster node ids to the
individual nodes.

> So indexes are locally builded, how does JackRabbit syncs them?

Every time some item has been modified, it's internal identifier is
logged in the journal. Another, "foreign" node, synchronizing with the
latest changes, will read that id, load the associated content and
fire an internal event that informs the search handler to re-index
this node. Of course, sharing the indexes themselves would be a better
approach, and this might come in a future release.

Dominique

>
> BR,
> Andrea -
>
>
>
> Dominique Pfister wrote:
> >
> > Hi Andrea,
> >
> > Since there have been lot of questions recently on how to setup
> > clustering and what options are available, I created a wiki page:
> >
> >   http://wiki.apache.org/jackrabbit/Clustering
> >
> > If you find that some important information is missing or wrong,
> > please do not hesitate to correct it!
> >
> > On 5/7/07, Andrea K. <akregar@alteanet.it> wrote:
> >> How is JCA resource (rar file) deployed in a cluster environment?
> >
> > JCA resource adapter will expect a repository.xml configuration file.
> > See the wiki page mentioned above to include cluster-specific
> > configuration details.
> >
> >> Indexes will be in a cluster-node shared filesystem folder?
> >
> > No, they won't. Indexes will be held and updated locally, in every
> > cluster node.
> >
> >> Can I use a datasource instead of
> >> OracleFileSystem/OraclePersistenceManager
> >> with Oracle DB?
> >
> > No, this isn't possible. In order to ensure consistency, Jackrabbit
> > must be in exclusive control of the persistence manager's JDBC
> > connection.
> >
> > Kind regards
> > Dominique
> >
> >>
> >> I'm using JackRabbit version 1.3 now, without clustering, and it works
> >> well.
> >>
> >> Sorry for the huge amount of questions!
> >>
> >> Thanks in advance.
> >> BR,
> >> Andrea -
> >> --
> >> View this message in context:
> >> http://www.nabble.com/Clustering-documentation-tf3703496.html#a10356658
> >> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
> >>
> >>
> >
> >
>
> --
> View this message in context: http://www.nabble.com/Clustering-documentation-tf3703496.html#a10359317
> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
>
>

Mime
View raw message