>I have smtpd_recipient_restrictions check_recipient_maps working and >dynamically updating this list/map for the domains that I host, but now >I (am just getting around to.. ) would like to perform recipient address >verification for the very few domains for which I relay mail, but don't >control/host the email servers for. These domains are in above list as >@domain.tld OK.
>So I would like to specify the short list of domains to >invoke the verify server. It seems like if I turn this on toward the end >of my smtpd_recipient_restrictions, that imgate will start a whole bunch >of probes to my servers ? As usual I am probably very confused by the >postfix manual. Please advise. one way is to create a restriction class for these domains, and process them separate for "main branch" of main.cf. /etc/postfix/reject_unverified_recipient.class contains: domain1.tld reject_unverified_recipient.class domain2.tld reject_unverified_recipient.class domain3.tld reject_unverified_recipient.class domain4.tld reject_unverified_recipient.class and then postmap it. it's a hash: file by default. in main.cf: #define the list of classes smtpd_restriction_classes = reject_unverified_recipient.class #define the restrictions to apply to this class reject_unverified_recipient.class = reject_unverified_recipient, . . . permit smtpd_recipient_restrcitions = . . . . hash:/etc/postfix/reject_unverified_recipient.class, reject_unlisted_recipient, . . . ===================== the logic: 1) if the @recipient.domain matches the reject_unverified_recipient.class, the msg gets branched into handling by "reject_unverified_recipient.class =", and DOESN'T RETURN to the main branch. So any restrictions that you want to apply to reject_unverified_recipient.class that come afterwards in main branch have to be also listed in the "reject_unverified_recipient.class =" branch. 2) if the @recipient.domain mismatches reject_unverified_recipient.class, the msg continues in the main branch. maybe there's a simpler way. I'm pooped. :) Len
