[169387] in North American Network Operators' Group

home help back first fref pref prev next nref lref last post

RE: Gmail throttling?

daemon@ATHENA.MIT.EDU (Nathan Anderson)
Sat Feb 22 07:56:02 2014

From: Nathan Anderson <nathana@fsr.com>
To: =?iso-8859-1?Q?=27Eduardo_A=2E_Su=E1rez=27?=
 <esuarez@fcaglp.fcaglp.unlp.edu.ar>, "'nanog@nanog.org'" <nanog@nanog.org>
Date: Sat, 22 Feb 2014 04:55:37 -0800
In-Reply-To: <20140221215857.t6rjafajs4ccww84@fcaglp.fcaglp.unlp.edu.ar>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org

On Friday, February 21, 2014 4:59 PM, Eduardo A. Su=E1rez <mailto:esuarez@f=
caglp.fcaglp.unlp.edu.ar> wrote:

> some of our users have forwarded the email to Gmail and Gmail now are
> complaining that this is bulk mail and delaying it.
>=20
> We have SPF, DKIM, DMARC, even SRS to try these things do not happen :(

Have you double-checked your setup to make sure it is performing SRS correc=
tly?  In my experience, Google is secretly blacklisting certain IPs for unk=
nown, unpublished reasons, and implementing SRS seems to be a surefire work=
around.

If you aren't on the secret blacklist, mail will still pass even if it fail=
s SPF, but once you are on the blacklist, mail that fails SPF (either softf=
ail or fail) will not be delivered.  If a user of yours is forwarding mail =
from your server to Gmail, the SPF check is not going to be against *your* =
SPF record, but against the original sender's SPF record, and so the check =
will fail (since the message looks like it is coming from you, and your MX =
won't be listed in the original sender's SPF record...thus, it will look li=
ke you are spoofing mail for the original sender).  Adding a valid SPF reco=
rd to your domain and then implementing SRS on your mail server should ensu=
re that all SPF checks pass, even for mail that your users are forwarding t=
o Gmail.

I wrote a post detailing my experience and findings: http://www.brokenbitst=
ream.com/gmail-spf-policy

--=20
Nathan Anderson
First Step Internet, LLC
nathana@fsr.com


home help back first fref pref prev next nref lref last post