jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Michael Harris" <michael.e.har...@gmail.com>
Subject Re: server modes/backups
Date Wed, 06 Aug 2008 17:11:48 GMT
Thomas thanx for replying.

So my organization has in place facilities to automatically back up oracle
dbs every day.  So the primary question

can we configure JR to store its content in oracle?

The next question is

if we do that will the oracle backup be sufficient to restore the system in
case of hardware failure?

On Wed, Aug 6, 2008 at 12:06 PM, Thomas Müller <thomas.mueller@day.com>wrote:

> Hi,
> > Basically, does it store its content and operate entirely on the
> > filesystem?
> Yes when using a file system based persistence manager. See also:
> http://wiki.apache.org/jackrabbit/PersistenceManagerFAQ - I wouldn't
> suggest to use file system persistence manager for production. If you
> use a database based persistence manager, this is also true if
> database content is also stored in the file system. This is the case
> when using the default persistence manager (Derby).
> > Or does it operate entirely in a database?
> > Or is this configurable?
> This is configurable in the repository.xml and workspace.xml.
> > What is derby being used for, and is it running entirely in memory?
> Derby is the default persistence manager, and it stores the data to disk.
> > Does it always need both a db and fs to operate?
> The search index always needs a file system if enabled. When using
> Jackrabbit for production, I suggest _not_ to use the file system
> persistence manager. But a database is not always required: The
> company I work for (Day, http://www.day.com) provides a transactional
> file based persistence manager (TarPersistenceManager) that stores
> everything to disk. This persistence manager is part of the CRX
> product (commercial). See also
> http://www.day.com/site/en/index/solutions/content-centric_infrastructure/content_repository.html
> - CRX is compatible to Jackrabbit.
> > For the purposes of server backup:
> >
> > If JR is storing its data on the filesytem, is it enough to zip the repo
> > home to back up the repository?  The idea being if there was some
> hardware
> > failure we could pull the zip and dump the contents into new machines
> repo
> > home.
> Yes, but you can't do that while the repository is running. If you
> need a failsafe solution, maybe you are interested in clustering?
> Jackrabbit supports clustering. Again the commercial CRX product from
> Day supports some more features in this area.
> > If JR is storing its data in a db, is it enough to back up the database
> > instance?
> You need to backup both the database and the file system.
> Regards,
> Thomas

Michael Harris

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message