[omniORB] Communications Failure on server after executes method

[Sai-Lai Lo]

dmorgen@alum.mit.edu (David Morgenlender) writes:

> I'm currently using omniORB 2.5.  I have a couple of methods running on
> the server (under ETS-Kernel).  The client is running on a different
> system (Win95). One method is working ok.  The other method seems to be 
> working ok, except a  "Communication failure" message is generated after 
> the method exits.  The message is generated in TFactory.cc in 
> tcpSocketWorker::_realRun().  It catches  a COMM_FAILURE exception,
> checks for a trace level >=3D 5, then generates the  message.

> The only significant difference between the calls to the 2
> methods, is  the problematic method has an unbounded sequence as an
> output parameter;  it consists of 500,000 ULong values;  the sequence 
> seems to be returned properly.
> 
> Do you have any idea what could be causing the problem & whether it's
> anything to be concerned about?

500,000 ULong = 2Mbytes. Your reply definitely exceeds the default limit on
the GIOP message size imposed by the ORB. Just increase the limit using
omniORB::MaxMessageSize(size_t newvalue) on *both* the client and the
server. See the user guide for details.

Regards,

Sai-Lai

-- 
Dr. Sai-Lai Lo                          |       Research Scientist
                                        |
E-mail:         S.Lo@orl.co.uk          |       Olivetti & Oracle Research Lab
                                        |       24a Trumpington Street
Tel:            +44 223 343000          |       Cambridge CB2 1QA
Fax:            +44 223 313542          |       ENGLAND