From: Rene van Hoek (reneactive8.nl)
Date: Tue Oct 17 2006 - 08:21:30 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Victor Duchovni wrote:
> On Tue, Oct 17, 2006 at 02:47:31PM +0200, Rene van Hoek wrote:
>
>> So, what will happen in this scenario:
>>
>> 1)
>> MAIL FROM: <joeexample.com> NOTIFY=FAILURE
>> delay_warning_time = 4;
>> message is delayed for more then four hours.
>>
>> The behavior I expect is that Postfix will not sent an delay message
>> because 'NOTIFY=FAILURE' is specified (sender did not request an delay
>> notification).
>
> When Postfix adds the string "DSN" to its EHLO response, its behaviour
> in cases covers by the relevant RFCs is (barring any unexpected
> implementation errors) as described in those RFCs. The rest follows...
>
Oke, so according too RFC3461, paragraph 5.2.5:

(a) If the NOTIFY parameter was supplied for a recipient and its
         value included the DELAY keyword, a "delayed" DSN MAY be issued.

    (b) If the NOTIFY parameter was not supplied for a recipient, a
         "delayed" DSN MAY be issued.

    (c) If the NOTIFY parameter was supplied which did not contain the
         DELAY keyword, a "delayed" DSN MUST NOT be issued.

So, If I 'read' the RFC correctly:

In case 'a' when delay_warning_time = 4, then Postfix will send a delay message.
Postfix will not send an delay message when delay_warning_time = 0.

In case 'b' when delay_warning_time = 4, then Postfix will send a delay message.
Postfix will not send an delay message when delay_warning_time = 0.

In case 'c' Postfix will not sent an delay message, although delay_warning_time = 4.

Maby I semm to be a little bit 'naggy', for that I appoligize. But I just want
to make sure that I interper the RFC correctly and understand the behaviour of
Postfix.

Rene van Hoek

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]