[JBoss Remoting Users] - J2EE remote client HTTP tunneling
by bwarren
I've got a swing client that's making EJB calls and listening to JMS topics in my JBoss app server. I get the server JNDI and do remote proxy lookups for EJBs, JMS Topics, and the JMS Connection Factory.
All of this works fine until somebody puts a firewall between my client and server and closes everything but port 80. I've been looking and I found quite a bit of possibilities but I didn't really see anything that looked like an exact fit. The closest I saw is JProxy, but testing reveals that it doesn't work. The proxy I got back still wanted to connect on a different port.
Is JBoss Remoting the way to go here? I could create a server object that proxies the calls to the various EJBs and to JMS and forward the results back to the client. Then tunnel all that over HTTP. My other option would be to convert my JMS code to use HTTP (not looked up from JNDI) and then create web services for my EJB calls. Is there something I'm missing? Maybe there's a way to configure the JNDI proxies to connect back with HTTP tunneling? We're also headed for clustering so I'll have that wrinkle in there soon. The non-firewall answer is to use HA-JNDI and everything else will work fine.
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4256981#4256981
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4256981
16 years, 9 months
[JBoss Cache Users] - 2 JBoss Cache questions
by userjboss99
Hello,
I'm a relatively new user to JBoss cache, and things generally seem to work well with JBoss cache while running inside of my application server. However, I am having 2 difficulties that I hope somebody out there can help me out with:
1) when using a JBoss cache instance in a standalone Java client outside of my JBoss application server (which running in a clustered mode), I seem to be unable to get it to join the application server cluster and have the caches talk to each other. I want to have my Java client application invalidate the application server's cache by using the INVALIDATION_ASYNC mode. I've started the JBoss server with the following arguments:
| ./run.sh -c all -b 10.1.0.22 -u 228.1.2.3 -g testcluster
|
My Java application code creates a cache like this:
| CacheFactory factory = DefaultCacheFactory.getInstance();
| Configuration config = new Configuration();
| config.setCacheMode(CacheMode.INVALIDATION_ASYNC);
| config.setClusterName("testcluster-query");
| jbCache = factory.createCache(config);
| jbCache.addCacheListener(new DebugCacheListener());
| jbCache.create();
| jbCache.start();
|
If I start up 2 of my Java standalone applications, I can see that these caches do indeed talk to each other (through the use of annotations on my DebugCacheListener such as @NodeInvalidated etc.)
2) in the same situation as #1 (using a cache instance in a standalone Java application), it seems that JBoss cache is creating lots of background threads that cause my application VM not to want to exit. Is there any way to have JBoss cache create these threads as daemon threads and not block the application VM from exiting automatically when my main() method finishes? I would like to avoid having to put System.exit() calls in my code to force a VM exit.
Thanks for any help you can provide!
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4256973#4256973
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4256973
16 years, 9 months
[jBPM Users] - Migrating custom workflow to jbpm
by p2auljr
I am working on a project to refactor a custom built workflow process (simple POJO design with state design pattern) to use jbpm. The one issue that I am concerned with is how to handle current processes that are not in a final state and that will need to be integrated into jbpm. I want to avoid having to add programming logic to traverse the workflow in order to put the process in a proper state within jbpm.
Is there a way to jump to a specific node when starting a new processInstance?
If not, I was thinking of having multiple "migration" transitions from the initial start node to handle processes that have started within the old workflow. The downside with that approach is it presents an untrue picture of the workflow.
Are there any best practices or recommendations for this?
Thank you.
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4256956#4256956
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4256956
16 years, 9 months