Dnia 23.06.2025 o godz. 15:33:06 sebastian via mailop pisze:
> They are implicitly trusted as you know they will not
> respond with fake DNS data in response to a malicious website.As opposed
> to someone linking in a 127.x.x.x record in their SPF which is impossible
> to anticipiate.

You are mentioning *website* and *SPF* together.

Do SPF records point to websites?
Do browsers look up SPF records in DNS?

What one has to do with the other?

The solution has already been mentioned here. Obviously a separate resolver
for a mail server (which is exempt from firewall), and separate one for
client machines running browsers, solves the problem and is easy to
implement. The best practice is that the mail server should have its own
resolver anyway.

In a resolver intended to use by a mail server *only*, you don't have to
worry about any "DNS rebinding" *at all*.

If you don't like that solution, pressure the firewall vendors to make them
fully stateful with regard to DNS and remember that the particular host name
appeared in a SPF record, so if a query is subsequently made for this host
name, the firewall should not mess with the reply. But this will probably
create other vulnerabilities that the attackers would want to exploit. The
two-resolver solution does not.
-- 
Regards,
   Jaroslaw Rafa
   [email protected]
--
"In a million years, when kids go to school, they're gonna know: once there
was a Hushpuppy, and she lived with her daddy in the Bathtub."
_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop

Reply via email to