Theo wrote:
> Abandoned_Trolley <
fr...@fred-smith.uk> wrote:
>> I am guessing that these dynamic IP changing tools manipulate the "A"
>> record so its possible that you could get it to work.
>
> There is a difference between getting it to work and getting people to
> receive your mail. If you run a mailserver from a dynamic IP pool:
>
> 1. Recipients are more likely to detect that and decide your messages are
> spam. Blackhole lists commonly used by mailservers record dynamic IP
> ranges.
>
> 2. You probably don't have the correct reverse DNS, which is another flag
> that recipients use to assume you're a spammer
>
> 3. You may end up on an IP that's been blacklisted because a previous user
> has been doing bad things. You can't practically fix that, because you
> might get shifted onto a different IP tomorrow.
This raises a more general question about IP addresses from a dynamic pool.
Any reputable mail server will check the IP address of a client wishing
to connect to it, irrespective of whether it is to send or receive; and
will block all connection attempts if that IP address is on a blacklist.
So anybody who gets a dynamic connection through CGNAT (so this includes
all mobile users, and I imagine many fibre users) will be unable to
connect to that mail server. The block is applied in the firewall
protecting the mail server, so it denies access to HTTPS traffic for
webmail, and POP3, IMAP, and SMTP traffic from a mail client.
So the person who uses a laptop or mobile in the office where it
connects via the LAN will be able to connect to the mail server, but if
that person goes on the road, or to another location where the
connection is through the 3G/4G/5G system, or (perhaps at a customer's
house) is through a domestic DSL service then the connection could fail.
Further, it will fail in a way that does not give the poor user any
idea as to why.
One possible solution is for the user to connect through a VPN. But
this then raises the possibility that any reputable VPN service provider
may also block traffic from IP addresses shown in blacklists.
Clearly a unique static IPv6 address issued to every internet connection
could solve the problem. This in principle would allow individuals
sending spam to be identified and blocked by their connection provider.
My guess is that would lead to a class of service providers who were
known for not checking what their customers send; thereby ensuring that
a whole population of users is blocked by reputable mail servers.
But generally, I can't see a sensible solution.
--
Graham J