This is an odd one:
<record>
<row>
<source_ip>65.20.0.12</source_ip>
<count>2</count>
<policy_evaluated>
<disposition>none</disposition>
<dkim>pass</dkim>
<spf>fail</spf>
</policy_evaluated>
</row>
<identifiers>
<header_from>mydomain.co.uk</header_from>
</identifiers>
<auth_results>
<dkim>
<domain>mydomain.co.uk</domain>
<result>pass</result>
</dkim>
<spf>
<domain>mydomain.co.uk</domain>
<result>fail</result>
</spf>
</auth_results>
</record>
I use a combination of my own servers and google to send mail. The source IP is not one of mine or one of Google's - how they hell can DKIM pass?
The reports come from Yahoo. The IP appears to be a mail server for btinternet.com managed by cpcloud.co.uk (Critical Path Inc.) - i know there was some oddness between them in the past with SPF etc. - could it be something to do with that?
The IP is only in the Yahoo DMARC reports. The dates I get the reports are 1 day ahead of emails that were sent to btinternet users.
Is it as simple as i'm sending an email to a bt account, it's getting forwarded/redirected/resent to yahoo and that's flagging the fail?