From: Roland Ramthun on
Hi all,

I obviously have an configuration error on my backup MX, but can't sort it out alone.

I use two mailservers, one primary MX and a secondary MX which relays messages to the primary.

A cron job on eara now generated a mail to root, which should be sent to mail(a)roland-ramthun.de.

This doesn't happen, instead postfix logs the fatal configuration error below.

I think the configuration error is misleading, what could the real problem be?
Can or should the message be fixed?

Thanks for your help,
Roland

###
Jul 1 17:30:00 eara postfix/master[32664]: daemon started -- version 2.6.6, configuration /etc/postfix
Jul 1 17:30:01 eara postfix/qmgr[32671]: 1B0CE781817: from=<root(a)eara.pxl.li>, size=2039, nrcpt=1 (queue active)
Jul 1 17:30:01 eara postfix/local[32674]: fatal: main.cf configuration error: mailbox_size_limit is smaller than message_size_limit
Jul 1 17:30:02 eara postfix/master[32664]: warning: process //usr/lib64/postfix/local pid 32674 exit status 1
Jul 1 17:30:02 eara postfix/master[32664]: warning: //usr/lib64/postfix/local: bad command startup -- throttling
###

###
eara ~ # mailq
-Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient-------
1B0CE781817* 2039 Tue Jun 29 05:01:31 root(a)eara.pxl.li
root(a)localhost.pxl.li

-- 2 Kbytes in 1 Request.
###

###
eara ~ # cat /etc/mail/aliases
# Basic system aliases -- these MUST be present.
MAILER-DAEMON: postmaster
postmaster: root

# General redirections for pseudo accounts.
adm: root
bin: root
daemon: root
exim: root
lp: root
mail: root
named: root
nobody: root
postfix: root

# Well-known aliases -- these should be filled in!
root: mail(a)roland-ramthun.de
# operator:

# Standard RFC2142 aliases
abuse: postmaster
ftp: root
hostmaster: root
news: usenet
noc: root
security: root
usenet: root
uucp: root
webmaster: root
www: webmaster
###

###
eara ~ # postconf -n
alias_database = hash:/etc/mail/aliases
alias_maps = hash:/etc/mail/aliases
anvil_rate_time_unit = 1h
anvil_status_update_time = 1h
command_directory = /usr/sbin
config_directory = /etc/postfix
daemon_directory = //usr/lib64/postfix
data_directory = /var/lib/postfix
html_directory = /usr/share/doc/postfix-2.6.6/html
inet_interfaces = all
mail_owner = postfix
mailbox_size_limit = 10737418240
mailq_path = /usr/bin/mailq
manpage_directory = /usr/share/man
maximal_backoff_time = 1h
maximal_queue_lifetime = 31d
message_size_limit = 134217728
minimal_backoff_time = 30m
mydestination = $myhostname, $mydomain, localhost, localhost.$mydomain
myhostname = eara.pxl.li
mynetworks_style = host
myorigin = $myhostname
newaliases_path = /usr/bin/newaliases
queue_directory = /var/spool/postfix
readme_directory = /usr/share/doc/postfix-2.6.6/readme
relay_domains = mysql:/etc/postfix/relay_domains.sql
relay_recipient_maps = mysql:/etc/postfix/relay_addresses.sql
sample_directory = /etc/postfix
sendmail_path = /usr/sbin/sendmail
setgid_group = postdrop
smtpd_banner = $myhostname ESMTP
smtpd_client_connection_count_limit = 3
smtpd_client_message_rate_limit = 200
smtpd_discard_ehlo_keywords = silent-discard, dsn
smtpd_helo_required = yes
smtpd_recipient_restrictions = reject_unknown_sender_domain, reject_unknown_recipient_domain, reject_non_fqdn_sender, reject_non_fqdn_recipient, permit_mynetworks, reject_unknown_client_hostname, sleep 1, reject_unauth_pipelining, reject_unauth_destination, check_policy_service inet:127.0.0.1:2501, permit
smtpd_sasl_auth_enable = no
###

From: Luigi Rosa on
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Roland Ramthun said the following on 01/07/10 19:39:


> I use two mailservers, one primary MX and a secondary MX which relays messages to the primary.

What's the point in setting a mailbox limit on a backup MX server?

If you set mailbox_size_limit to zero, what happens?


Ciao,
luigi

- --
/
+--[Luigi Rosa]--
\

Superior ability breeds superior ambition.
--Spock, "Space Seed"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkws16UACgkQ3kWu7Tfl6ZTKKACcD91Zp7NSaVzIFLDq57MAHamy
JDQAn2Ms1IgCxFqz15Q9/vNAPcxQ9QO1
=bO/x
-----END PGP SIGNATURE-----

From: Brian Evans - Postfix List on
On 7/1/2010 1:39 PM, Roland Ramthun wrote:
> Hi all,
>
> I obviously have an configuration error on my backup MX, but can't sort it out alone.
>
> I use two mailservers, one primary MX and a secondary MX which relays messages to the primary.
>
> A cron job on eara now generated a mail to root, which should be sent to mail(a)roland-ramthun.de.
>
> This doesn't happen, instead postfix logs the fatal configuration error below.
>
> I think the configuration error is misleading, what could the real problem be?
> Can or should the message be fixed?
>

The problem is the configuration, see below:
> ###
> Jul 1 17:30:00 eara postfix/master[32664]: daemon started -- version 2.6.6, configuration /etc/postfix
> Jul 1 17:30:01 eara postfix/qmgr[32671]: 1B0CE781817: from=<root(a)eara.pxl.li>, size=2039, nrcpt=1 (queue active)
> Jul 1 17:30:01 eara postfix/local[32674]: fatal: main.cf configuration error: mailbox_size_limit is smaller than message_size_limit
> Jul 1 17:30:02 eara postfix/master[32664]: warning: process //usr/lib64/postfix/local pid 32674 exit status 1
> Jul 1 17:30:02 eara postfix/master[32664]: warning: //usr/lib64/postfix/local: bad command startup -- throttling
> ###
> mailbox_size_limit = 10737418240
> message_size_limit = 134217728
global/mail_params.h
563:extern int var_mailbox_limit;

Both of these parameters are signed integers as of Postfix 2.6.5 (and
possibly later versions)
Once you exceed 2,147,483,647; you get overflows.

Maybe Wieste should investigate changing the mailbox limit to a long (if
he hasn't already) as disk space becomes cheaper all of the time.

Brian

From: Roland Ramthun on
On Thu, 01 Jul 2010 14:03:23 -0400
Brian Evans - Postfix List <grknight(a)scent-team.com> wrote:

> global/mail_params.h
> 563:extern int var_mailbox_limit;
>
> Both of these parameters are signed integers as of Postfix 2.6.5 (and
> possibly later versions)
> Once you exceed 2,147,483,647; you get overflows.

This seems to be the case: Changing the mailbox limit to 0, as Luigi pointed out, or <2,147,483,647 both solved the problem.

As the limit is given in bytes this means you can't have mailboxes larger than ~2048MB.
If this is a design decision, please consider to document it e.g. in the postconf man page (at least that's where I looked for info).

Thanks for your quick help.

Regards,
Roland

From: Roland Ramthun on
On Thu, 01 Jul 2010 20:00:09 +0200
Luigi Rosa <lists(a)luigirosa.com> wrote:

> What's the point in setting a mailbox limit on a backup MX server?
>
> If you set mailbox_size_limit to zero, what happens?

The configuration was partly copied from an old machine, this setting doesn't make sense in this setup.
Changing the value to 0 solves the problem, see my other mail.

Thanks for your help,
Roland