Re: Using ServiceStarter to launch a "proxy is service" service

>> Okay well I think we need to back up a little first:
>> 
>> (1)     What governs the decision to deploy everything remotely or
>> locally?

Well... each service controls a device attached to a COM port. An application can control more than ...  say 20 devices. So we can have 
        Reason 1: to distribute processing load
        Reason 2: port limitations
        Reason 3: Serial cable limitations
	
 
>> (2)     Why when in "local mode" do we want to avoid the overhead?
>>

In order to clarify my questions I presented a simplified version of my application. My clients are themselves services which communicate with other services in my application. Thus in a transaction, involving n services, I could have a significant gain of performance just in avoiding the rmi overhead of calls. Assuming that I can have 100 transactions/minute...´


Regards,
Francisco Mourão



> -----Original Message-----
> From: Dan Creswell [mailto:[email protected]]
> Sent: terça-feira, 30 de Janeiro de 2007 15:03
> To: [email protected]
> Subject: Re: Using ServiceStarter to launch a "proxy is service" service
> 
> Hi Francisco,
> 
> Francisco Dário Mourão wrote:
> >>> Remember also the most common structure for a service is to have
> >>> a proxy and one or more back-ends and thus it might not be
> >>> entirely surprising to find certain things (like
> >>> com.sun.jini.start) favour that model over others such as the
> >>> much rarer proxy only service.
> >
> > I thought about this "proxy is service" model because of the
> > following:
> >
> > - In some installations my application launches n clients in a
> > computer and m services in a remote computer. It's a real distributed
> > environment. In this case scenario I use the RMI proxy model (with
> > service proxies and service backends).
> >
> > - Although, at some installations everything (clients and services)
> > runs in the same computer. For the moment I'm still using the RMI
> > proxy model for this scenario too. But I'm trying to remove the rmi
> > overhead when everything runs in the same computer. Is this valid
> > question? Or is it an absurd?
> >
> 
> It's a valid question, no question ;)
> 
> > So, I want my application to have a parameter in each client saying
> > "how" it will use the service: (a) using it remotely through a proxy
> > (RMI proxy model) (b) using it locally ("proxy is service" model)
> >
> 
> Okay well I think we need to back up a little first:
> 
> (1)     What governs the decision to deploy everything remotely or
> locally?
> 
> (2)     Why when in "local mode" do we want to avoid the overhead?
> 
> My gut instinct is that it's wrong for the client to hint on the "how"
> of the service because it's the service that dictates the "how"
> normally.  Thus I'm attempting to ascertain what it is in your
> environment that really dictates this need and then fashion an
> appropriate answer to your implementation question.
> 
> > Basically, I want to write a service which is prepared to be used
> > with both models...
> >
> > Is this thought valid? Is there an easy way to do this?
> >
> 
> We'll find out!
> 
> Best wishes,
> 
> Dan.
> 
> --------------------------------------------------------------------------
> Getting Started:     http://www.jini.org/wiki/Category:Getting_Started
> Community Web Site:  http://jini.org
> jini-users Archive:  http://archives.java.sun.com/archives/jini-users.html
> Unsubscribing:       email "signoff JINI-USERS"  to [email protected]

--------------------------------------------------------------------------
Getting Started:     http://www.jini.org/wiki/Category:Getting_Started
Community Web Site:  http://jini.org
jini-users Archive:  http://archives.java.sun.com/archives/jini-users.html
Unsubscribing:       email "signoff JINI-USERS"  to [email protected]