On Wed, May 11, 2011 at 11:26 AM, Simon Laws <simonsl...@googlemail.com> wrote: > Quick question. Ant, I notice that you've been adding more information > to the endpoint (domain?) registry. Also there are some operations in > the domain node that use some of that information to present > information, such as the domain composite. Is that information > available from a domain node without starting a composite in that > node? > > I'm asking as I'd like to start up a process to look into the domain. > Re-using that code makes sense but I don't want to have to start a > composite just to look at what's already in the domain. >
That is exactly what i'm presently working on to try to get going better, at the moment some things work and others don't so its a bit hit and miss, mainly beacuse we didn't/don't have all the information in the registry, it should be going better soon as i get it finished. So for example if you start two shells, eg with one with mvn tuscany:shell -DdomainURI=uri:default, and the other in the helloworld sample with mvn tuscany:run -DdomainURI=uri:default, then in both shells if you type "services" you should see the helloworld service endpoint, if you type "installed" you should see the installed helloworld contribution. You can't yet use the invoke command in either because in the remote shell it doesn't yet have the service interface so can't create the proxy, but if you where doing it programatically with then API calls then you can invoke the service in the remote node. What i'm trying to finish now is to have all the node api calls and shell commands all reflected everywhere across the distributed domain. So on one node you can install a contribution, on another see that contribution and start composites in it etc. ...ant
