On Feb 25, 2015, at 5:56 PM, Noel Jones <njo...@megan.vbhcs.org> wrote:
> On 2/25/2015 4:53 PM, Research wrote: >> Hello, >> >> I have recently begun deploying Postfix on a web server. >> >> Postfix is configured to handle the e-mail for the web server domain (i.e.: >> receives e-mail for example.com), and then has virtual tables configured to >> route that mail to Gmail accounts. I have mappings for all RFC required >> e-mail addresses (i.e.: postmas...@example.com, ab...@example.com), and can >> verify that mail sent to those addresses does go to the recipients on Gmail >> successfully. I have also verified that I am not an open relay. >> >> In my syslog, I can see an e-mail transaction of the following form >> (truncated): >> >> Feb 25 04:48:14 example postfix/qmgr[16092]: BBCE57FCD4: from=<>, >> size=20118, nrcpt=1 (queue active)... >> >> In this case, a sender has attempted to deliver a message with no sender >> (“from”), information. As far as I can tell, that is not-RFC compliant and >> is most likely spam. >> >> Is there a way via: smtpd_sender_restrictions in main.cf that I can block >> this behaviour, or is there another method of doing so ? >> >> Thank you. >> > > > The null sender is used for RFC compliant bounces or non-delivery > notices and MUST NOT be blocked. > > It's OK to block a particular message if the client or content > violates some local policy of yours, but the null sender MUST NOT be > used as blocking criteria. > > > -- Noel Jones Hi Noel, Thank you for your prompt response. I was unaware of bounce and non-delivery notices using null sender addresses and would have attempted to block this to my servers’ detriment. I will use content filtering to examine these addresses and *NOT* block solely on null addresses.
