On 2022-11-22 at 00:05:06 UTC-0500 (Mon, 21 Nov 2022 21:05:06 -0800)
Sean Kamath via mailop <kam...@moltingpenguin.com>
is rumored to have said:
On Nov 21, 2022, at 18:29, John Levine via mailop <mailop@mailop.org>
wrote:
I understand why that's the conventional wisdom, but I also don't
understand why, if all the resources are on the same LAN as the name
servers, the conventional wisdom would apply.
It’s nice to have mail queued instead of rejected? Seems to me, if
the NS servers are failing you get a reject for a domain that
“doesn’t exist”, while if you have active NS records and an
unreachable MX you get mail queued until the machine(s) come(s) back.
Generally speaking, SERVFAIL responses (as in this case) or
non-answering nameservers are not treated by MTAs as equivalent to
explicit NXDOMAIN responses, but instead as transient problems meriting
later retrial.
--
Bill Cole
b...@scconsult.com or billc...@apache.org
(AKA @grumpybozo and many *@billmail.scconsult.com addresses)
Not Currently Available For Hire
_______________________________________________
mailop mailing list
mailop@mailop.org
https://list.mailop.org/listinfo/mailop
