1. Here's is a fairly busy IMGate using reject_unverified_recipient (I 
would prefer he use to_relay_recipients for these volumes), and only 4 of 
the bounces are by his Imail:

Per-Hour Traffic Summary
     time          received  delivered   deferred    bounced     rejected
     --------------------------------------------------------------------
     0000-0100        7834        542         31         15       9854
     0100-0200        7463        435          8         13       9951
     0200-0300        7193        517         15         11       9053
     0300-0400        7588        510          9          5       9598
     0400-0500        7342        456         10          9       9155
     0500-0600        7824        653         14          5      10065
     0600-0700        7087        504         13          4       9096
     0700-0800        8108        920         19         10      11110
     0800-0900        7552       1323          9         14       9098
     0900-1000        7719       1074         57          8       9413
     1000-1100        7697       1095         46          8       9171
     1100-1200        9011       1077         23          6      11201
     1200-1300        7478        978         43         12       9295
     1300-1400           0          0          0          0          0

2. and his harvested _dict file plays a typically big role in total rejects:

    3323 RBL bulk.rhs.mailpolice.com
    3327 DNS nxdomain for MTA PTR hostname (forged @sender.domain)
    4938 ACL SAV: unverifiable sender address
    9833 ACL mta_clients_bw
   14254 SMTP Exceeded Hard Error Limit after DATA
   14617 RBL cbl.abuseat.org
   16512 ACL RAV: undeliverable recipient address  <<<<<<<<<<<<<<
   20951 SMTP Exceeded Hard Error Limit after RCPT
   21060 ACL mta_clients_dict                      <<<<<<<<<<<<<<
   26197 ACL helo hostname contains an IP
============================
  152237 TOTAL

Usually, for a day, the RAV rejects will be bigger than the _dict rejects.

Len


Reply via email to