On 5/18/2015 12:32 PM, aleph de wrote: > I'm relaying mail received from the 'net at one Postfix server on a VPS, at a > 'real' IP address, to another Postfix server in my office at an 'internal' IP > address, over a VPN link. > > The way I currently do this is the VPS Postfix listens on a real IP, then > relays 'from' an IP bound to an internal address assigned to a dummy intfc on > the VPS. > > inet_interfaces = 1.1.1.10, 1.1.1.11 > smtp_bind_address = 10.1.0.17 > > This works with mail getting received and delivered all the way through to > the office like I intend. > > I'm going through all the headers at each stage line by line to make sure > it's working. > > I notice that on mail successfully received on the office server, the headers > include > > Received: from relaymx.MYDOMAIN.com (unknown [10.1.0.17]) > (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) > (Client CN "relaymx.MYDOMAIN.com", Issuer "MYDOMAIN_CA" (verified OK)) > by officemx.MYDOMAIN.com (Postfix) with ESMTPS id B6C33101D6A > for <postmas...@mydomain2.com>; Mon, 18 May 2015 10:02:16 -0700 (PDT) > > Is that > > "unknown [10.1.0.17]" > > an actual problem? Or just something to take note of?
It can cause problems in some configurations, but since your mail is flowing correctly, you can ignore it. The safe choice is to fix it before it creates a problem (which may be never). > > I guess it's becuase there's no RDNS on that internal IP? Yes, that's the cause. > > If that's true, do I have to figure out how/where to assign that on the VPS, > or can I set that 'statically' inside my Postfix configuration? > Adding an appropriate entry in the /etc/hosts file is the usual quick fix. Or add the 10.in-addr-arpa zone to your local DNS. -- Noel Jones
