well the definition of cdi beans is pretty explicit: each beans are a
singleton in a particular scope. So basically no pool with defaults scopes
(you can still imagine your own weird scope ;).

that said, this thread is interesting but the title doesn't match anymore
the content, if you want to continue maybe we should open another thread,
no?

*Romain Manni-Bucau*
*Twitter: @rmannibucau <https://twitter.com/rmannibucau>*
*Blog: **http://rmannibucau.wordpress.com/*<http://rmannibucau.wordpress.com/>
*LinkedIn: **http://fr.linkedin.com/in/rmannibucau*
*Github: https://github.com/rmannibucau*



2013/3/14 James Green <[email protected]>

> By comparison, are beans held in a CDI scope managed in a pool with min/max
> boundaries?
>
> I've been trying to avoid using EJBs - probably due to seeing them being
> registered at boot and my thinking that it's probably a waste of RAM when
> they are idle. Right now I have @Stateless on my JAX-RS and JAX-WS classes
> but use CDI on the backend.
>
> A repercussion of this was that @Dependant beans were being reused by the
> JAX-RS classes across invocations which I hadn't appreciated would be the
> case. I tried switching to @RequestScoped but something blew up. Elsewhere
> an @RequestScoped bean with a constructor that had @Inject parameter also
> blew up. I had to remove these quickly to get a demo up and running again
> but it revealed a level of interaction complexity I had not anticipated
> from skim reading what I found in examples on the web.
>
>
> On 14 March 2013 03:59, Anthony Fryer <[email protected]> wrote:
>
> > +1 for Stateless Session Beans.  I've never had an issue with them and
> find
> > them a good way to control resource utilization.  The only things to keep
> > in
> > mind is what your min and max pool sizes are for the beans.  If you make
> > that match up with the number of threads, then you shouldn't have any
> > threads blocking waiting for a free bean.  I usually set min and max to
> the
> > same value since i don't see much value in having a pool shrink or grow.
> > Also good to keep in mind if you inject other EJBs into an EJB, then its
> > the
> > EJB with the smallest pool size that could cause a thread to block.
> >
> >
> >
> > --
> > View this message in context:
> >
> http://openejb.979440.n4.nabble.com/ConcurrentAccessTimeoutException-tp4661288p4661516.html
> > Sent from the OpenEJB User mailing list archive at Nabble.com.
> >
>

Reply via email to