Hi, Am 16.02.2016 um 23:33 schrieb Christian Schneider: > Yes. To have double the performance in a binary transport is also what I > would expect. > I also think the performance you measured for CXF should be fine. > > I wonder why ECF and Amdatu would be so much slower. Maybe there is just > something wrong with the setup. After looking into the implementation, Amdatu does not use connection pooling for the HTTP connection. So each service invocation creates a new connection to the other side which is directly closed after that. That probably explains its poor results.
For ECF I think it's best when we share our simple test code and for Scott to have a look at it :) Greetings -Sascha-