From: Dean Strik (no email)
Date: Fri Aug 01 2003 - 08:36:23 EDT
wrote:
> Hello
>
> The proposed solution is not working.
> The mailer doesn't take the second MX record when I shut down the internal-1
> server
It should. For example:
sphinx# host -tmx stack.nl
stack.nl mail is handled (pri=10) by skynet.stack.nl
stack.nl mail is handled (pri=10) by hexagon.stack.nl
stack.nl mail is handled (pri=100) by kweetal.tue.nl
stack.nl mail is handled (pri=100) by mailhost.tue.nl
stack.nl mail is handled (pri=10) by vaak.stack.nl
Bringing a few MX-es down:
skynet# postfix stop
vaak# postfix stop
Mailing:
sphinx# echo test | mail -s test
Maillog:
Aug 1 14:31:30 sphinx postfix/qmgr[59823]: 5B33022840:
from=<>, size=305, nrcpt=1 (queue active)
Aug 1 14:31:30 sphinx postfix/smtp[80469]: connect to
vaak.stack.nl[2001:610:1108:5010:248:54ff:fe6b:2cec]: Connection refused
(port 25)
Aug 1 14:31:31 sphinx postfix/smtp[80469]: 5B33022840:
to=<>, relay=hexagon.stack.nl[2001:610:1108:5010::144],
delay=1, status=sent (250 Ok: queued as 0E6DF1C7B)
If one MX is down, it should try the other. Please show some maillogs
similar to the above to see whether it tries to connect to the second
host.
Please also give a postconf -n and the exact line in your transport maps.
-- Dean C. Strik Eindhoven University of Technology | | http://www.ipnet6.org/ "This isn't right. This isn't even wrong." -- Wolfgang Pauli
|
|
|