Alright, I will wait a bit longer, but I also feel this is right strategy
On Mon, May 27, 2013 at 4:37 PM, Daniel Bevenius
>For now, we could use CDI events in combination with the
EJB annotation (for publishing and receiving).
Sounds like a good option to me.
On 27 May 2013 08:09, Matthias Wessendorf <matzew(a)apache.org> wrote:
> On Sun, May 26, 2013 at 6:45 AM, Douglas Campos <qmx(a)qmx.me> wrote:
>> On Thu, May 23, 2013 at 02:28:47PM -0400, Jay Balunas wrote:
>> > One question that came up around the JMS stuff is how will it work on
>> > top of plain old Tomcat? Will we be including the extra libs for that
>> > case? Or perhaps that functionality is only available if JMS is part
>> > of the container?
>> That's exactly why I was unhappy with this change =/
> fair point, right now JMS is not used for much.
> Once a request hits the server (e.g. request against our HTTP sender
> API), we produce a JMS message and send a "JOB Submitted" 200 HTTP
> The consumer for the JMS message does all the work that _may_ take
> longer, and is blocking IO (->JPA).
> E.g. it looks up the MobileVariantInances and is responsible for
> delivering the messages to the actual push networks (GCM/APNs).
> I think, that we can remove the JMS part, at least for now. (e.g. for the
> early August release). However, I am sure, that at some point messaging
> will helpful for scaling/throughput.
> For now, we could use CDI events in combination with the @Asynchronous
> EJB annotation (for publishing and receiving).
> would that be OK ?
>> aerogear-dev mailing list
> Matthias Wessendorf
> blog: http://matthiaswessendorf.wordpress.com/
> sessions: http://www.slideshare.net/mwessendorf
> twitter: http://twitter.com/mwessendorf
> aerogear-dev mailing list
aerogear-dev mailing list