...
So...I guess I have a couple of questions.
1. What difference would the SME server stmp proxy make to not be able to deliver the email?
2. could it be the SME server has cached the address of the old destination and just not using the correct address?
...
If the client's domain or one of the client's MX servers is defined in some way on your SME server you might get odd behavior.
Here are some commands to run on the SME server, and some info on what they might tell you:
db domains show |grep -i domain.com- if the client's domain is defined and you have NOT also configured specific mail routing rules, the SME server will try to accept all email for the domain.
- If the client's domain is defined, it might also be configured to use a specific dns server -- you'll need to check that dns server for incorrect information.
- If the domain is defined, you could have an old/invalid setting configured for 'MailServer'
dig domain.com -t mx
#
# then, for each host listed as an mx server for domain.com, do some checking on your sme:
db domains show |grep -i <last part of mx hostname>
db hosts show |grep <hostname>
- If you have the client's primary mx hostname defined on your SME then your SME will use the locally defined address and not the publicized address.
db accounts show |grep pseudonym- If you have a pseudonym defined for person@domain.com then the pseudonym will (probably) override any other settings on the SME.
Finally, you can find out from /var/log/qmail/current what your SME server is actually trying to do when it attempts to deliver email to domain.com