G'day,

I've got a bit of a problem, I've rtfm'ed the bits I believe I need to
have read but I'm still stuck (or not understanding the bits I've
read...).

We've got postfix 2.5.5 running on Debian with majordomo installed and
working successfully.  I've tested majordomo and it's working fine (in
my tests and with about 100 other lists), however we have a largish
mailing list ~55-60,000 subscribers (~3000 local) which is causing us
some grief...

For local users Postfix delivers via lmtp to a remote cyrus mail store,
it will occasionally recieve a temporary 452 over quota and the
occasional 550 no such user if the mailbox has been recently deleted,
and should deliver non local recipients via smtp as normal.

This all works well for most of the lists we have, however on the
largish list (without fail in my testing today) I get for the first
three deliveries (which all happen to be local): 550 Mailbox unknown,
250 Successful delivery, followed by a 452 Over quota. and then postfix
immediately sticks it in the deferred queue and then attempts to deliver
to the same 3 recipients 10 minutes later, without attempting to deliver
to any other addresses.  I can see other lists where postfix is
successfully reciving a 452 over quota but still going and delivering to
the rest of the recipients on the list.


Doing a /usr/sbin/sendmail -bv <list-address>
It goes through the motions and appears to lists many, many addresses
but itself gets placed in the deferred queue at somepoint.

grepping through the logs for 'qmgr.*(panic|fatal|error|warning):' comes
up whith nothing.  And no other related warnings or errors, the machine
itself is not stressed in any way for cpu/io/memory, there's only about
70 messages in the mail queue at present.

I've tried tinkering with default_recipient_limit,
lmtp_initial_destination_concurrency and initial_destination_concurrency
without success.

In the mail queue it shows up:
        C5012280E0     6390 Wed Apr 22 16:56:29 
owner-commun...@infoxchange.net.au
        (host 192.168.210.99[192.168.210.99] said: 452 4.2.2 Over quota (in 
reply to RCPT TO command))
                 over-quota-username-repla...@infoxchange.net.au
                 community-mj-l...@infoxchange.net.au

The majordomo alias file is:
        community:          "|majordomo_wrapper resend -l community 
community-mj-list"
        community-mj-list:  :include:/home/majordomo/lists/community
        owner-community:    communitybounce
        community-request:  "|majordomo_wrapper request-answer community"
        community-approval: listmaster

This is a fairly newly put in service machine, this mailing list has
been running quite well on a very old postfix install (postfix 1.1!) for
years, the major difference being local lmtp socket delivery with 5xx
errors when a user is over quota.

What am I doing wrong?  I'm hoping someone will be able to tell me what
I'm doing wrong or failing to understand :-)

Steve.
===============
postconf -n
address_verify_map = btree:/var/lib/postfix/verify
address_verify_negative_expire_time = 3h
address_verify_negative_refresh_time = 1h
alias_database = hash:/etc/postfix/aliases                        
hash:/etc/postfix/majordomo
alias_maps = hash:/etc/postfix/aliases                  
hash:/etc/postfix/majordomo
append_dot_mydomain = no
biff = no
body_checks = pcre:/etc/postfix/body_checks
broken_sasl_auth_clients = yes
config_directory = /etc/postfix
content_filter = smtp-amavis:[127.0.0.1]:10024
debug_peer_level = 2
default_process_limit = 150
default_transport = smtp
fast_flush_domains = $relay_domains
header_checks = pcre:/etc/postfix/header_checks
html_directory = /usr/share/doc/postfix/html
inet_interfaces = all
inet_protocols = ipv4
local_command_shell = /usr/sbin/smrsh -c
local_recipient_maps = $alias_maps, hash:/etc/postfix/lu
mailbox_size_limit = 0
mailbox_transport = lmtp:192.168.210.99
message_size_limit = 10240000
mydestination = mail3.infoxchange.net.au infoxchange.net.au 
localhost.localdomain localhost
mydomain = infoxchange.net.au
mynetworks = /etc/postfix/mynetworks
myorigin = $mydomain
readme_directory = /usr/share/doc/postfix
recipient_canonical_maps = regexp:/etc/postfix/recipient_canonical
recipient_delimiter = +
relay_domains = $mydestination                    /etc/postfix/relay-domains
relay_transport = smtp
relayhost = 
sender_canonical_maps = regexp:/etc/postfix/sender_canonical
smtpd_banner = $myhostname ESMTP mail3.infoxchange.net.au
smtpd_client_restrictions = reject_unauth_pipelining smtpd_client_misbehaving 
permit
smtpd_helo_restrictions = permit
smtpd_recipient_restrictions = virus_check mj_protect 
  reject_non_fqdn_sender reject_non_fqdn_recipient reject_unknown_sender_domain
  quota_check permit_mynetworks
  permit_sasl_authenticated       reject_unauth_destination
  reject_unverified_recipient     good_nets       good_hosts
  check_sender_good       spam_friends    bad_nets        bad_hosts
  check_sender_bad        check_rbls      policy_other    policy_greylist
smtpd_restriction_classes = bad_hosts   bad_nets
  check_non_sending_addresses     check_rbls      check_sender_bad
  check_sender_good       good_hosts      good_nets       mj_protect
  non_sending_addresses   no_outside_forge_admin  others_abused_by_us
  policy_greylist policy_other    reject_dsn      reject_dsn_to
  reject_internal_admin   smtpd_client_misbehaving        spam_friends
  virus_check     virus_reject    quota_check
smtpd_sasl_auth_enable = yes
smtpd_sasl_exceptions_networks = $mynetworks
smtpd_sasl_local_domain = $mydomain
smtpd_sasl_path = smtpd
smtpd_sasl_security_options = noanonymous
smtpd_sasl_type = cyrus
smtpd_sender_restrictions = check_non_sending_addresses no_outside_forge_admin  
reject_dsn      permit
soft_bounce = no
transport_maps = hash:/etc/postfix/transport
unverified_recipient_reject_code = 450

-- 
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?

Reply via email to