jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefan Guggisberg <stefan.guggisb...@gmail.com>
Subject Re: mongodb or couchdb?
Date Sun, 21 Nov 2010 15:53:09 GMT
On Fri, Nov 19, 2010 at 8:41 PM, John Langley
<John.Langley@mathworks.com> wrote:
> Has anyone recently looked at creating a persistence manager based on either
> mongodb or couchdb?

what kind of improvements would you expect from a mongodb or couchdb
persistence manager?

if you're after raw performance, i'm rather skeptical. the ideal
persistance store is IMO
a simple but highly optimized in-proc key-value store (e.g. something
like an embedded
java berkeley db). embedded derby, although a full rdbms, comes IMO
pretty close.

i doubt that a standalone engine (couch/mongo) will be able to provide
a better performance
than an embedded engine, but i might be wrong ;-)

BTW: implementing a couchdb/mongodb pm should be straight forward.
the PersistenceManager interface is pretty simple [1].


[1] http://jackrabbit.apache.org/api/2.1/org/apache/jackrabbit/core/persistence/PersistenceManager.html

> I did see the discussion about Unified Persistence and jr3. But it doesn’t
> seem like anyone’s considering doing this for jackrabbit 2.x Is that right?
> Are there reasons not to?
> Thanks in advance for any pointers
> -          Langley

View raw message