On Fri, 2011-05-13 at 00:55 +0530, Asankha C. Perera wrote: > On 05/13/2011 12:30 AM, Oleg Kalnichevski wrote: > > It is hard to tell without seeing a wire log or a code snippet. Is there > > any chance you could reproduce the problem with a self contained > > application? > Hi Oleg > > Its a bit late over here.. so let me attach the raw mock echo server I > am using to test this. When sending the "close-write" header on a > request, it writes one byte back and closes the socket.. > > However I assume my earlier analysis on the Java Channel.transferFrom() > maybe more useful, as I suspect the issue lies there. If so, we could > expose the fact that a partial response has been received, so that > higher level code can use that instead for error handling? > > regards > asankha >
Asankha I am also puzzled by the code snipped you posted in your previous message. The way they handle IOExceptions makes no sense to me at all. I am also unable to reproduce the problem using the test app, as it is not clear to me how the client side is expected to interact with it. Worse, currently I do not see how this problem correlates with my latest changes. Let us pick it up tomorrow. Oleg --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
