On Wed, 21 Oct 2009, johnea wrote:

> OT?: I commented out the: #  -o milter_macro_daemon_name=ORIGINATING
>      since this wasn't a listed value for this parameter in the docs,
>      and I wasn't sure what that line was doing.

If you do not use milters, then you can safely ignore this.

> The main.cf remained unchanged with these restrictions:
> 
> atom# postconf -n | grep _restrictions
> smtpd_helo_restrictions = permit_mynetworks,
>                           reject_non_fqdn_helo_hostname,
>                           reject_invalid_helo_hostname,
>                           reject_unknown_helo_hostname,
>                           permit

You can remove that trailing permit.

> smtpd_recipient_restrictions = permit_sasl_authenticated,
>                                permit_mynetworks,
>                                reject_invalid_hostname,
>                                reject_non_fqdn_hostname,
>                                reject_non_fqdn_sender,
>                                reject_non_fqdn_recipient,
>                                reject_unknown_sender_domain,
>                                reject_unknown_recipient_domain,
>                                reject_unknown_client_hostname,
>                                reject_unauth_destination,
>                                check_policy_service unix:private/policy
> smtpd_sender_restrictions = permit_mynetworks,
>                             reject_unauthenticated_sender_login_mismatch
> 
> 
> Another related but somewhat OT question: Since the users are now migrated
> to exclusively using the submission port; Should I remove the 3
> permit_mynetworks and the permit_sasl_authenticated?

Leave them there, unless you want SASL authenticated clients to be
subject to the reject_foo_bar restrictions that follow.  Remember, since
you only override your main.cf smtpd_(helo|client)_restrictions in
master.cf, all clients (via the submission service or otherwise) are
subject to smtpd_recipient_restrictions. 

> If the permit_* statements are removed, should reject_unauth_destination be
> moved to the top of the smtpd_recipient_restrictions list?

I think it is fine as is.

-- 
Sahil Tandon <sa...@tandon.net>

Reply via email to