AT&T's Blacklist site suggests one option of correcting mail server issues is:

Reconfigured server to ignore NDRs to forged addresses/domains

However, NDRs are returned to the sender by their own server's mail daemon. So, it seems that one's mail server could never SPAM another mail server by sending NDRs.

Another indirect form of spamming is backscatter. Could this be what AT&T is referring to? If so, how could one go about securing (or confirming) their mail server to address this issue?

Current setup:

  • APF blocking previously malicious IPs
  • Rejecting mail from 4 RBLs
  • Rejecting SPAM via SpamAssassin SPAM score
  • Rejecting SPAM via SpamAssassin domain policies
  • Disabled catchalls for all client accounts
  • Discarding (instead of bouncing) SpamAssasssin SPAM for large volume accounts
  • Registered server on abuse networks of major mail hosts
  • Monitoring RBLs for listing notices

In short, to prevent backscatter you configure your mailserver to reject an undesired e-mail message during the SMTP transaction, rather than accepting the message first and then rejecting it at a latter stage.
By not accepting the message it remains the responsibility of the sending mailserver and it is not your mailserver but rather the sending mailserver that has to generate the Non-Delivery Report.

Alternatively you configure your mailserver/spamfilter/antivirus software to not reject undesired mail, bit to simply discard it silently or quarantaine it.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.