For our MDaemon hosting we can easily bind each postoffice to an IP address. So all spam filter domains are on one IP and all other domains are on a separate IP, only two IPs needed. Our perimeter firewall restricts incoming SMTP to only the spamhero.com MTA ip address ranges for the spam filter domains. This way MDaemon will only accept email for the spam filter domains from spamhero.
For MailEnable we had to end up with an advanced filter configuration like so:
Code: Select all
If CriteriaMet([ME_IPADDRESS],"108.60.195.*") OR _
CriteriaMet([ME_IPADDRESS],"208.53.48.*") OR _
CriteriaMet([ME_IPADDRESS],"X.X.X.*") OR _
CriteriaMet([ME_FROM],"*@domain.com") OR _
CriteriaMet([ME_HEADERS_CONTAIN],"*MailEnable WebMail*") OR _
[ME_IPADDRESS_WHITELISTED] = 1 Then
FilterResult = 0
This allows spamhero IP addresses, our own IP address range, from the same domain or from webmail or a whitelisted IP. We needed all of these parameters to handle the various options since there didn't seem to be one way to handle them all:
- Spamhero IP range
- Internal user to another internal user over SMTP (does not leave the server so does not hit spam filter)
- Internal user to another via WebMail
From what I understand we can not bind multiple domains to multiple IP addresses. So I am at a loss as to how to move forward with any scalability. We can not rely on these kinds of filters to handle this as it gets messy and prone to human error. Being able to flip a switch by moving a domain to one IP or another is easy and clear. Is there no other way to handle this scenario?