[Replies on list, please]
On Wed, 2007-01-31 at 16:50 +0500, Boris Kovalenko wrote:
> What bits? :) I have no specific manualroutes, and just one default
> dnslookup router
> dnslookup:
> driver = dnslookup
> domains = ! +local_domains
> transport = remote_smtp
> ignore_target_hosts = 0.0.0.0 : 127.0.0.0/8
> no_more
>
> remote_smtp:
> driver = smtp
> interface = 80.64.16.18
Which is good. Now we know there's nothing odd going on - prior to this
we had to assume.
> Transport: relay.grosbein.pp.ru [213.184.64.166] error 60: Operation
> timed out
> first failed: 31-Jan-2007 08:09:51
> last tried: 31-Jan-2007 13:40:34
> next try at: 31-Jan-2007 15:55:34
OK, all looks perfectly reasonable.
> I know, but there is no fallback to the second MX :(
But this is odd.
As far as I know this *should* fall back to the second MX, however...
there has been no IP-level rejection (connection refused, ICMP
network/destination/host/port unreachable or similar). If I try to
telnet to the host, my session sits there until telnet decides it's been
too long and it wants a cup of coffee, then gives up. There's nothing
from the remote end telling us that the server's dead, just silence.
I'd hazard a guess that this is where your problem lies - this isn't
really a "failure" that you could call by a name, just... silence. Now
that *should* result in trying the second MX.
What do you see if you grab one of the message-IDs you have queued and
do:
exim -d -M MESSAGEID
?
Graeme
