Hi Charles,
yes, agree, we can have both Tomcat and Jetty, each one dedicated to
some functions.
Regards
JB
On 04/21/2010 05:54 PM, Charles Moulliard wrote:
Hi JB,
Many thanks for your feedback. As jetty is an important component in
camel and servicemix, I'm not quite sure that it makes sense to remove
it for the moment while camel-tomcat and servicemix-http are not
available.
We can also design the product using jetty and Apache HTTP Server:
http://wiki.eclipse.org/Jetty/Tutorial/Apache
So we don't need to design new components. Apache HTTP Server is use
in plenty projects where Websphere AS is deployed and WAR file can be
deployed in Jetty :
http://www.enavigo.com/2008/08/29/deploying-a-web-application-to-jetty/.
As a wrapper exist, it can be started as a Service on Windows, ...
Kind regards,
Charles Moulliard
Senior Enterprise Architect (J2EE, .NET, SOA)
Apache Camel Committer
*******************************************************************
- Blog : http://cmoulliard.blogspot.com
- Twitter : http://twitter.com/cmoulliard
- Linkedlin : http://www.linkedin.com/in/charlesmoulliard
On Wed, Apr 21, 2010 at 5:25 PM, Jean-Baptiste Onofré<[email protected]> wrote:
Hi Charles,
as already discussed together, the idea looks very interesting to me.
By packaging Tomcat, Camel, ODE and JBI components in ServiceMix (and of
course working on the documentation :)), we can provide a wide scope and
powerful ESB. Combining with the OSGi/Karaf give us a very flexible and
modern platform.
Tomcat can be an interesting alternative to Jetty to in components (there is
some work to migrate).
Definetely +1 for me.
Regards
JB
Charles Moulliard wrote:
Hi,
ServiceMix EAI. The name should be probably changed but the idea
having an Enterprise Integration Server is to combine the strength of
the ServiceMix ESB server and Routing Engine with Tomcat to offer a
platform where J2EE applications can also be deployed.
Why : Many clients are still designing and developing their projects
as J2EE applications which are deployed as WAR or EAR in an
Application Server. People in charge of the infrastructure have skills
and competences to manage such J2EE applications. When they have to
manage a new kind of server like ServiceMix, they are more reluctant
as they have less skills and return of experience. But if we can
propose a packaged version of Tomcat where ServiceMix is already
deployed, they will accept. We can also say the same thing for the
development team because the clients have invest since more than a
decade in Java/Web developers + Spring recently.
The other advantage that I see also concerning this product is that we
can propose to the clients an environment where there applications can
be easily load-balanced, services could be deployed as bundles and
accessed from J2EE application using OSGI service (like IBM WebSphere
does - see Aries project for that and WAB). In fact, we can propose an
Open Source SOA solution leveraging of the stregnth of Application
Server World, OSGI modularity, Messaging Bus, Routing, ... Combining
with REST/WebService, ... who can inter-operate with any other
existing
system. Loadbalancing is a key success factor in the architecture when
we have to process thousands of requests quickly and when we have to
distribute the work load between different servers (= cloud
computing).
What do you think about that ?
Charles Moulliard
Senior Enterprise Architect (J2EE, .NET, SOA)
Apache Camel Committer
*******************************************************************
- Blog : http://cmoulliard.blogspot.com
- Twitter : http://twitter.com/cmoulliard
- Linkedlin : http://www.linkedin.com/in/charlesmoulliard
