postfix-users October 2010 archive
Main Archive Page > Month Archives  > postfix-users archives
postfix-users: my backup mail server isn't delivering by itself

my backup mail server isn't delivering by itself

From: Jay G. Scott <gl_at_nospam>
Date: Wed Oct 13 2010 - 16:18:43 GMT


I should know the answer to this but I don't. Very embarrassing.

The problem is my inside the firewall, backup mail server.
mxhost2 (ns6) in the dig output below.
# dig -t MX

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> -t MX
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 8190
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 3, ADDITIONAL: 5


;; ANSWER SECTION: 21600 IN MX 15 21600 IN MX 10

;; AUTHORITY SECTION: 21600 IN NS 21600 IN NS 21600 IN NS

;; ADDITIONAL SECTION: 21600 IN A 21600 IN A 21600 IN A 21600 IN A 21600 IN A

;; Query time: 0 msec
;; WHEN: Wed Oct 13 11:02:46 2010
;; MSG SIZE rcvd: 219

Here's the, trimmed to save bandwidth.

myhostname =
mydomain =
myorigin = $mydomain
inet_interfaces = all
mydestination =
local_recipient_maps =
relay_domains =
relayhost =

relay_recipient_maps = hash:/VOLUMES/postfix/etc/pfknown_users
smtp_generic_maps = hash:/VOLUMES/postfix/etc/pfgenerics
alias_maps = hash:/VOLUMES/postfix/etc/pfmalias, dbm:/VOLUMES/postfix/etc/aliases
transport_maps = hash:/VOLUMES/postfix/etc/pftransport

local_transport = error:local mail delivery is disabled

Here's the question:

This backup machine, IMHO, ought to be able to deliver mail to internal
machines without going through the primary, internal mail server
(mxhost1). It isn't. Mail leaving my site from ns6 should go straight
to ns4, and not touch mxhost1. It isn't. It seems ns6/mxhost2 can't
do anything without mxhost1. Why?

My answer is that (from the dig output) mxhost1 is lower cost than
mxhost2 everything goes to mxhost2. That explanation isn't entirely
satisfactory even to me, because even outbound mail from ns6 goes
through mxhost1 even though the relayhost is set to ns4.

So, my first question is, to be explicit: why is mxhost1 getting
involved? Why isn't mxhost2/ns6 taking care of this itself?

Second question: what needs to change so that mxhost2/ns6 does start
doing things without bothering mxhost1?

(I sent test emails from mxhost2/ns6 to inside and outside addresses
and the headers show they all go through mxhost1.)


-- Jay Scott 512-835-3553 Head of Sun Support, Sr. System Administrator Applied Research Labs, Computer Science Div. S224 University of Texas at Austin