jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jukka Zitting" <jukka.zitt...@gmail.com>
Subject Re: Use of Jackrabbit in commercial product and scalability numbers for Jackrabbit w/w/o RMI
Date Mon, 14 May 2007 17:19:02 GMT

On 5/14/07, ruchi goel <Ruchi.Goel@sun.com> wrote:
> Jukka Zitting wrote:
> > What sort of scalability are you interested in? Number of nodes, total
> > amount of content, number of (concurrent) users, perfromance under
> > load, ...
> Will be interested in as many I can get .

We unfortunately don't have very organized scalability/performance
test suites in place to give very specific numbers, but I've seen
successful load tests that go up to tens of millions of nodes and
terabytes of content per workspace.

Jackrabbit also performs quite well on concurrent read access thanks
to the aggressive read caching that takes place at the item state
manager level.

However, under the item state managers all persistent content accesses
are currently serialized (see JCR-314), which causes a bottleneck for
heavily concurrent write loads and read loads with widely distributed
access patterns.


Jukka Zitting

View raw message