BEA WebLogic Server 8.1SP2 Solaris 8

-Customer is using -Dcom.sun.CORBA.ORBServerPort flag to bind to a port
-Once the remote application is rebooted the invoking the method causes
the error

-Dcom.sun.CORBA.ORBServerPort setting is causing the error. This setting
was used by BEA to bind rmi-iiop objects to certain port.

java.rmi.MarshalException: ; nested exception is:
        org.omg.CORBA.BAD_PARAM: vmcid: OMG minor code: 23 completed: Maybe
        at weblogic.iiop.OutboundRequestImpl.sendReceive(OutboundRequestImpl.java:187)
        at weblogic.iiop.IIOPRemoteRef.invokeInternal(IIOPRemoteRef.java:209)
        at weblogic.iiop.IIOPRemoteRef.invoke(IIOPRemoteRef.java:153)
        at com.Servigistics.AutoPilot.IAutoPilotServer_IIOP_WLStub.doMethod2(Unknown Source)
        at com.Servigistics.AutoPilot.AutoPilotClient.doMethod2(AutoPilotClient.java:85)
        at jsp_servlet.__test._jspService(__test.java:127)
        at weblogic.servlet.jsp.JspBase.service(JspBase.java:33)
        at weblogic.servlet.internal.ServletStubImpl$ServletInvocationAction.run(ServletStubImpl.java:971)
        at weblogic.servlet.internal.ServletStubImpl.invokeServlet(ServletStubImpl.java:402)




BEA engg., has looked into this and have noted the following and
subsequently concluded the WebLogic JNDI is not caching the IOR, and
hence they want to hear from Sun JDK team about this issue at this
point. This issue has caused a major pain point with one of our
customers and we would need a resolution ASAP.

The bad param exception is coming because of the following exception
thrown by the jdk orb on the corba server side.

com.sun.corba.se.internal.iiop.ServerRequestImpl(Thread[RequestProcessor[0],5,ORB
ThreadGroup]): Sending SystemException: org.omg.CORBA.BAD_PARAM: vmcid: OMG minor code: 23 completed: Maybe
        at com.sun.corba.se.internal.iiop.IIOPInputStream.createWCharBTCConverter(IIOPInputStream.java:216)
        at com.sun.corba.se.internal.iiop.CDRInputStream_1_0.getWCharConverter(CDRInputStream_1_0.java:2309)
        at com.sun.corba.se.internal.iiop.CDRInputStream_1_2.read_wstring(CDRInputStream_1_2.java:68)
        at com.sun.corba.se.internal.iiop.CDRInputStream_1_0.read_value(CDRInputStream_1_0.java:1050)
        at com.sun.corba.se.internal.iiop.CDRInputStream.read_value(CDRInputStream.java:293)
        at com.Servigistics.AutoPilot._Object1IIOP_Tie._invoke(_Object1IIOP_Tie.java:72)
        at com.sun.corba.se.internal.corba.ServerDelegate.dispatch(ServerDelegate.java:353)
        at com.sun.corba.se.internal.iiop.ORB.process(ORB.java:280)
        at com.sun.corba.se.internal.iiop.RequestProcessor.process(RequestProcessor.java:81)
        at com.sun.corba.se.internal.orbutil.ThreadPool$PooledThread.run(ThreadPool.java:106)

###@###.### 2005-05-11 22:08:50 GMT

Problem happens when a corba obejct-to-name binding is restarted and the client has a stale IOR from an earlier connection request. Client side is 3rd party orb and server side is Sun's orb. The problem is subsequent requests of the client, to invoke remote methods(RMI-IIOP) fail, though the client has, now, correct OR/newly fetched IOR.


###@###.### 2005-07-19 18:17:41 GMT