Your idea of wrapping the interface returned by the registry was a stroke of brilliance! And I'm pleased to say it was successful.
I'm now working to resolve another small hiccup. The IoC registry can be configured to provide the service as either a singleton or a new instance for each new thread. I chose the latter. Socket connections are established without problem, but if Tomcat subsequently calls @OnError and/or @OnClose on a connection, the IoC registry provides a new instance (not good). This suggests that the subsequent method call on the socket comes via a new thread. So bit of a mismatch there between sessions and threads. I may have to move to a singleton service, meaning no longer being able to persist connection endpoints to a static map with a simple: connections.put(this). Thanks again, Chris. --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
