Hi all,
this is a request for help with diagnosing log viewer messages from the j2ee server. I have a c++ client that accesses an EJB over IIOP. I recently switched the ORB I was using on the client end, and now the client cannot successfully connect to the EJB.
The following is a summary of the j2ee log viewer messages for both the new (bad) orb and old (good) orb when the c++ client invokes the EJB. With the bad orb, it looks like to me that it (the j2ee server, or to be more exact the iiop dispatcher) hits an out of memory exception in the middle of unmarshalling the first String parameter. The good orb goes through the exact same steps, and has the same index values, but doesn't hit the memory problem, and the invocation works fine. What could account for this difference ?
(note: the messages are ordered most recent first)
BAD ORB messages:
Throwable in invoke: java.lang.OutOfMemoryError
endTag: -1
skip start block...
endBlockIndex: 2147483392
chunked: false
START BLOCK: 107
REP ID: IDL:omg.org/CORBA/WStringValue:1.0
VALUE TAG: 2147483392 POS: 68
READ VALUE at:
READ VALUE at: 64
Connection of the servant object to the ORB passed successfully
Request from 10.48.160.56:53007
InQMy ORB, multiple server threads.
Request from 10.48.160.56:53007
Request from 10.48.160.56:53007
GOOD ORB messages:
READ VALUE RETURNS: class java.lang.String
skip start block...
endBlockIndex: 2147483392
chunked: false
START BLOCK: 90
STORE INDIRECTION AT: 68
END TAG
chunked: false
READ END TAG: 90
endBlockIndex: 2147483392
END BLOCK: 90
endTag: -1
skip start block...
endBlockIndex: 2147483392
chunked: false
START BLOCK: 107
REP ID: IDL:omg.org/CORBA/WStringValue:1.0
VALUE TAG: 2147483392 POS: 68
READ VALUE at:
READ VALUE at: 64
Connection of the servant object to the ORB passed successfully
Request from 10.48.160.56:53007
InQMy ORB, multiple server threads.
Request from 10.48.160.56:53007
Many thanks,
Colm.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.