Corba problem, migration from JDK 1.4 to 1.5
Kuba Witczak
I faced some errors with CORBA, after performing migration from JDK 1.4.1 to
JDK 1.5.0_02
on Solaris 8 2/02 (SPARC version).
The problem occurs in a web application launched on Tomcat 5.5.7, if JDK
1.5.0_02 is used and trying to connect with CORBA server.
If using JDK 1.4.1 the CORBA stuff works fine.
The second problem is that I don't have direct access to the CORBA server,
and cannot test it by myself.
That's why I'm preparing ideas and small command line test applications, so
I could test many possible solutions over a short period of time.
I present my current ideas below the problem description, if you have some
other ideas please post them. TIA
I spent few days on digging in java.sun.com bug database and google archive,
but haven't found similar case.
I isolated the problem code and found out, that Tomcat cannot cause the
error, because i receive the same error while invoking test class from
command line.
The client code uses IOR file for obtaining the object reference and default
JDK Orb.
The stacktrace(see below) goes through new package in JDK 1.5
com.sun.corba.se.impl.* and java.lang.Class which was rewritten in JDK 1.5
in order to use generics. Is it possible that the bug lies somewhere in this
classes?
Is this a known issue? Does anyone have a workaround?
Have you seen some similiar behavior associated with migration to JDK 1.5.0
?
My current ideas include:
+ using external ORB implementation (JacORB)
+ applying Solaris patches recommended by site
http://sunsolve.sun.com/pub-cgi/show.pl?target=patches/J2SE
+ checking if the same client using JDK 1.5.0_05 works properly
eventhough the specification from
http://java.sun.com/j2se/1.5.0/ReleaseNotes.html doesn't explicitly
mention improvements in Corba field, i believe some of low-level bugs,
fixed in latest updates, might have an impact on Corba behaviour.
What do you think about them?
I'm running out of ideas, on debugging this stuff, any help or suggestion
could be helpful.
Thanks In Advance, Kuba
PS. I attach the received stack trace below:
org.omg.CORBA.BAD_OPERATION: vmcid: 0x0 minor code: 0 completed: Maybe
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native
Method)
at
sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:39)
at
sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:27)
at java.lang.reflect.Constructor.newInstance(Constructor.java:494)
at java.lang.Class.newInstance0(Class.java:350)
at java.lang.Class.newInstance(Class.java:303)
at
com.sun.corba.se.impl.protocol.giopmsgheaders.MessageBase.getSystemException(MessageBase.java:897)
at
com.sun.corba.se.impl.protocol.giopmsgheaders.ReplyMessage_1_2.getSystemException(ReplyMessage_1_2.java:99)
at
com.sun.corba.se.impl.protocol.CorbaMessageMediatorImpl.getSystemExceptionReply(CorbaMessageMediatorImpl.java:572)
at
com.sun.corba.se.impl.protocol.CorbaClientRequestDispatcherImpl.processResponse(CorbaClientRequestDispatcherImpl.java:430)
at
com.sun.corba.se.impl.protocol.CorbaClientRequestDispatcherImpl.marshalingComplete(CorbaClientRequestDispatcherImpl.java:326)
at
com.sun.corba.se.impl.protocol.CorbaClientDelegateImpl.invoke(CorbaClientDelegateImpl.java:129)
at org.omg.CORBA.portable.ObjectImpl._invoke(ObjectImpl.java:457)
at
generated.from.idl.package._ServerFactoryStub.FindServer(_ServerFactoryStub.java:25)
...