Looking at the error-report, it would seem that the machine (or host)
'filter.maildefender' (which is presumably a local LAN-only hostname,
given that it has no TLD suffix of any kind) has somehow failed to
complete a DNS lookup for
MYDOMAIN.demon.co.uk.
That would indicate that there was either a transient failure whereby
IntY were unable to reach their own DNS cache of Demon's domain-data
(please god they DO have their own local copy of Demon's DNS?) or they
were unable to reach Demon's authoritative DNS servers upon which to
look up your address.
Either way, it smacks of infrastructure weakness between IntY and Demon
at some level. At least with the old punt mail system, it (a) had access
to its own copy of the Demon DNS zone data at its fingertips (and
several backup servers), all 'on the local net' so to speak and (b)
didn't have to reach out across the dirty big internet to talk to the
DNS servers to query local mailbox and domain info, and (c) would simply
hang on a lot longer until it got a valid DNS result for a local user,
before giving up with a hard-failure error.
MS's own Exchange blurb says this about a 5.4.4 extended SMTP error (at
http://support.microsoft.com/kb/284204 )
===QUOTE===
Numeric Code: 5.4.4
First Available: Exchange 2000 Service Pack 1
Possible Cause: No route to message, next hop not found. You have set up
a Routing Group topology, but there is no Routing Group Connector set up
between the Routing Groups.
Troubleshooting: Add or configure your Routing Group Connector between
Routing Groups.
===ENDQUOTE===
IMHO, it's a self-inflicted loss of DNS, by virtue of them having borked
their Exchange/MSSMTP routing config. Are they really up to handling an
entire ISP's primary mailstore, if this kind of fubar can happen? Even
temporarily?
--
Neil