Johannes Zillmann wrote:
--- Johannes Zillmann <[EMAIL PROTECTED]> schrieb:
I tried also the messageSent() variant, but without
success...
I'm sorry, i made a mistake.
Now i tried Niklas 'messageSent() variant' and it
seems to work very good!
Do one of you if it could be a problem if the reader
is much slower then writer ?
That means if writer has successfully witten 100MB but
read just reads it step by step.
Thank you so much for helping!
Johannes
So you've tried both approaches and the synchronous one (using
WriteReuqest.join()) doesn't work while the asynchronous (using
messageSent()) does? That's very interesting!
I'm sensing a bug here. I will write a test tonight to see if I can find
anything.
No, it shouldn't matter if the reader is slower than the writer. Your
OS's write buffer will be filled and MINA won't get notified until
there's room in the buffer for more data again. And your handler or
filter won't get notified (through messageSent()) until MINA is notified
by the OS. The same SHOULD also be true for WriteRequests when using join().
/Niklas
