Hi Viktor, Terrific advice. Much appreciated.
Cheers. Andy On Mon, Jun 2, 2014 at 11:35 AM, Viktor Dukhovni <postfix-us...@dukhovni.org > wrote: > On Mon, Jun 02, 2014 at 11:26:47AM +1000, tswmmeejsdad . wrote: > > > On Mon, Jun 2, 2014 at 10:19 AM, Viktor Dukhovni < > postfix-us...@dukhovni.org > > wrote: > > > > > On Mon, Jun 02, 2014 at 08:35:25AM +1000, tswmmeejsdad . wrote: > > > > > > > May 19 16:06:50 smtp01 postfix/smtpd[21940]: timeout after > END-OF-MESSAGE > > > > from unknown[10.84.72.227] > > > > > > The SMTP client neither closes the TCP socket nor sends QUIT after > > > completing a message delivery. Fix the client. > > > > Thanks for the info Viktor. That sounds like a rather bold statement and > I > > don't think customers at times are willing to concede that it's something > > on their end causing it. > > No need to be shy about stating the truth. > > > Is there anything else that we could do further to prove to the customer > > that it was their application that as you have said is not closing the > TCP > > socket nor sending QUIT after completing a message delivery? > > A packet capture. Your log excepts are much too short to conclude > that the unterminated connection is from the customer. All I said > was that that SMTP client neither closed the connection, nor sent > QUIT. That SMTP client might be some middle box between you and > the customer, or even some proxy on your own server. > > You need to look at the complete logging to determine whether the > log messages in question are a result of a direct connection from > the customer, or mediated in some way. > > -- > Viktor. >
