[omniORB] Lost one-way calls

[Nick Reeves]

Could this be the case where the server side of the connection does an
idle timeout at just the wrong time, and hence drops the message on
the floor.

I have always though that because of this problem the default idle
timeout should be significantly higher on the server side than the
client side. At the present they are the same.

Duncan Grisby writes:
 > On Thursday 22 April, "Neykov, Plamen" wrote:
 > 
 > > We are developing a distributed calculation engine and the problem occurs
 > > during the assignment of calculation tasks. 
 > > A coordination server, which knows the whole task is assigning pieces of
 > > work to the calculation servers which are ready with their prior task. The
 > > assignment is realized through a one-way call. Some of the calc-servers
 > > after completing the task don't receive the call from the coordination
 > > server, but the coordinator doesn't get any exception after the one-way call
 > > (we would expect COMM_FAILURE to be thrown). As far as we can see, some of
 > > the one-way calls are simply lost ...
 > 
 > Oneway calls cannot throw any sort of exceptions, so if they fail you
 > never get to know about it. In fact, there are no guarantees that a
 > oneway call will actually happen. That said, omniORB does try quite
 > hard to send oneway calls -- can you give more details about what you
 > are doing when this problem occurs?  Also, is there a good reason why
 > you're using oneways rather than normal operations?
 > 
 > Cheers,
 > 
 > Duncan.
 > 
 > -- 
 >  -- Duncan Grisby  \  Research Engineer  --
 >   -- AT&T Laboratories Cambridge          --
 >    -- http://www.uk.research.att.com/~dpg1 --

-- 
Nick Reeves