posted 21 years ago
I haven't articulated my problem well again. There's nothing wrong with my server - if an instance of RMI server is running that knows about the DbEngine stub location, is running, then the server-side I wrote runs fine. Similarly, if the server-side is running fine, the client can hit it fine - all of this works from jarred up solutions from the command line.
What doesn't work from the command line, is the RMI server itself. I can run the RMI server from my IDE, and I can pass it a location of the stub as a classpath, then it runs, and I can run my server side. But from the command line, when I call , I can't pass it the location of the stub, which is now in the jar file, (or if I can, I don't know how). As a consequence, when I run the server-side that I wrote, it throws a huge nested exception. In pertinent part it says:
I hope I'm clear that the only difference between success and failure here is that, when I'm successful, the call to rmiregistry.exe knows about the classpath to the stub (because I'm making that call from inside the ide where it has that classpath). In the failed case, the call to rmiregistry has no path to the stub, so, when the server tries to bootstrap to it, it fails. In both the sucessful and unsuccessful case, the code I wrote is in the same 2 jar files. So this isn't a failure of the client to see the server. Its a failure of the registry to honor the call from the server, apparently only because the registry doesn't know where the stub is located.
Sun Certified Java Developer<br />Sun Certified Java Programmer