In an attempt to better fight spam, I've configured Postfix like this:
smtpd_client_restrictions =
check_client_access hash:/etc/postfix/client_whitelist
reject_unknown_client
smtpd_recipient_restrictions =
permit_mynetworks,
permit_sasl_authenticated,
reject_unauth_destination
reject_non_fqdn_sender,
reject_non_fqdn_recipient,
reject_unknown_sender_domain,
reject_unknown_recipient_domain,
reject_unauth_destination,
reject_unauth_pipelining,
reject_invalid_hostname,
reject_rbl_client zen.spamhaus.org,
reject_rbl_client bl.spamcop.net,
reject_rbl_client l2.apews.org,
permit
This config has reduced a lot of spam, and with the whitelist I was able to deal with some misconfigured but legitimate servers in order to accept them. However, I'm wondering if this particular whitelisting config will be able to do the job if some other legitimate servers found themselves blacklisted in one of these RBLs. Am I missing something?