On May 16, 2007, at 11:12 AM, jian wu wrote:
If I understand correctly, Mina treats the message A sent out and Message B received as two independent events, you might want to consider tracking "sending message A" in the filterWrite() of your last IoFilter before Mina actually writes out message A.
yup. could also put a callback on the WriteFuture for message A as well.sounds like the remote end is just replying *really fast* before you have a chance to receive the "sent event"
perhaps you could re-architect how you are doing things so that you advance your state machine right before you write the message, so you are prepared for an immediate reply.. and then if the write fails, back it up and re-try the write. might side-step the issue nicely.
-pete -- (peter.royal|osi)@pobox.com - http://fotap.org/~osi -- [EMAIL PROTECTED] - http://fotap.org/~osi
smime.p7s
Description: S/MIME cryptographic signature
