geronimo-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Russell E Glaue <>
Subject Re: How to deploy web application to one port
Date Wed, 17 Dec 2008 16:33:21 GMT
Russell E Glaue wrote:
> Okay, so I have to deploy a second Jetty container inside Geronimo, so Geronimo
> would have two different Jetty containers, one serving web applications on each
> of the two ports.
> So I would basically copy the Jetty plugin to a new plugin, compile it, and
> place it in the repository.
> So now we have JettyContainer/JettyWebContainer in Geronimo.
> After doing what is suggested, I would have something like
> JettyContainer2/JettyWebContainer2 - it would be a whole new module with the
> same code - is this correct?

The solution for this is discussed in the thread
"feature? ContainerGroup" 12/05/2008 09:46 AM.

The solution requires deploying a new jetty container via a gbean.
All that is required to be in the gbean is the plan.xml file which configures
the new instance of the Jetty container.

> I have an additional question.
> In a previous thread on 2008-12-03, I ask on the subject "How to set up logging
> for a connector". Can I also perform this same suggestion for
> org.apache.geronimo.jetty6.requestlog.NCSARequestLog ?
> I would assume if they use the same NCSARequestLog gbean that both containers
> would put their logs in the same log file.
> So... since I cannot change the log file name and location (as I asked in this
> other thread), I assume I also need to deploy a new NCSARequestLog gbean - is
> this correct?

Also answered in the thread "feature? ContainerGroup"
12/05/2008 09:46 AM, this requires deploying a new NCSARequestLog gbean. Again
the gbean only contains a plan.xml to configure the new instance of NCSARequestLog.

I have entered a feature request into the Jetty project's issue system.

This feature request is to implement a new RequestLog that does more than
file-based logging. The suggestion has been to use slf4j in the new
implementation, which in turn will support sending logs to Socket, Syslog, and
will in turn support log4j.


> -RG
> David Jencks wrote:
>> On Dec 4, 2008, at 10:59 PM, wrote:
>>> You can create two geronim instances in one server:instance A in http
>>> port
>>> 2150, instance B in port 2051, then deploy hello_1 to instance A,
>>> hello_2 to
>>> instance B.
>> I think creating two entire server instances is a bit extreme and e.g.
>> if both web apps are connecting to the same ejb app might not work as
>> efficiently as possible.
>> However, AFAIK you have to set up a second jetty server in your geronimo
>> instance.  Basically, copy the plugins/jetty plan (from source), call it
>> something else, edit it (e.g. the ports), and deploy it.  I'd copy the
>> maven project and build a plugin, but you can deploy the plan separately
>> if you want.  You can also include all the gbeans in your geronimo plan
>> for your web app.  You need to indicate in the geronimo plan which web
>> server you want the app to be deployed on.
>> There are micro instructions for geronimo-tomcat here related to a
>> sample app:
>> If you come up with something relevant for jetty and want to help us
>> with our docs and samples that would be great!
>> I think I recall talking with Greg about filtering web apps per port in
>> jetty but I don't think you can actually do it right now.  I think he
>> said it would be very easy to add though.... maybe we can get it into
>> jetty 7.
>> thanks
>> david jencks
>>> Russell E Glaue wrote:
>>>> Using geronimo-jetty 2.1.3
>>>> How do I deploy a web application to one and only one single port?
>>>> Let's say I have configured two Listen Sockets:
>>>> 1)
>>>> 2)
>>>> If I simply deploy a web application to Geronimo as in:
>>>> % deploy HelloWorld.war
>>>> The sample application HellowWorld is accessible on both ports as in:
>>>> 1) - SUCCESS!!!
>>>> 2) - SUCCESS!!!
>>>> Now, I can modify WelloWorld/WEB-INF/geronimo-web.xml with the
>>>> <virtual-host></virtual-host> attribute, but this restrict
>>>> web
>>>> application to be served when the web client makes a HTTP 1.1 call to
>>>> Geronimo
>>>> for "" GET /hello
>>>> So with using this <virtual-host>, the HellowWorld is not available
>>>> either
>>>> port when accessed as <server-ip>:<port-number>
>>>> So, how do I deploy a web application so that it is available on one
>>>> port,
>>>> but
>>>> not another port, yet both ports running under one geronimo instance?
>>>> I want to deploy two web applications, each one accessible on a
>>>> different
>>>> port.
>>>> If I do something like:
>>>> % deploy HelloWorld_1.war
>>>> % deploy HelloWorld_2.war
>>>> I want the results to be like:
>>>> 1) - SUCCESS!!!
>>>> 2) - Fail
>>>> 3) - Fail
>>>> 4) - SUCCESS!!!
>>>> How do I do this?
>>>> -RG
>>> -- 
>>> View this message in context:
>>> Sent from the Apache Geronimo - Users mailing list archive at

View raw message