> On Nov. 9, 2016, 9:50 p.m., Benjamin Mahler wrote: > > What you have looks good, I would just suggest that you also cover the case > > where the request body does not complete to EOF, but fails. In this case > > the server should not receive the request at all (if > > !RouteOptions.streaming) or should see the request.pipe fail (if > > RouteOptions.streaming). > > > > I looked through the existing tests for the response streaming, they are > > pretty lacking, you can blame me for that :) > > Also, they are a bit inconsistent with how you added the test here, so > > hopefully at some point we can circle back and clean up the existing tests.
hmm, `socket.shutdown()` only closes the receive end of the socket. It's not quite possible to add the test for the latter case as the server never realizes that the client went away. We need to fix that before a test can be added. Would file an issue and assign to me. - Anand ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/53490/#review155502 ----------------------------------------------------------- On Nov. 4, 2016, 5:55 p.m., Anand Mazumdar wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/53490/ > ----------------------------------------------------------- > > (Updated Nov. 4, 2016, 5:55 p.m.) > > > Review request for mesos and Benjamin Mahler. > > > Bugs: MESOS-6466 > https://issues.apache.org/jira/browse/MESOS-6466 > > > Repository: mesos > > > Description > ------- > > Added a test for request streaming via the connection abstraction. > > > Diffs > ----- > > 3rdparty/libprocess/src/tests/http_tests.cpp > 533104c93dd1eaf67bf3752163d2e0cad090078d > > Diff: https://reviews.apache.org/r/53490/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Anand Mazumdar > >
