[GLLUG] Postgrey holding list

Marco van Beek mvanbeek at supporting-role.co.uk
Tue Mar 3 14:55:10 UTC 2020


I'm sorry, I really thing you are over-engineering the problem. In as
simple a language that I can, this is what happens in postfix: 

 	* An SMTP connecting is made. Postfix accepts the initial HELO, Mail
>From and Rcpt To.
 	* Postfix supplies this information to PostGrey
 	* Postgrey does a lookup of the sending server (I think it does it by
host name from the EHLO but it might be by IP address), and the from
address.
 	* If it finds no entries, it adds one with the current time stamp, and
rejects the SMTP connection with a "temporarily unavailable" message.
 	* If it finds an entry, it checks the stored time stamp against the
current time. If it is over the delay setting, it allows it and updates
the timestamp.
 	* Somewhere along the way there is a cleanup task that removes old
entries.

That's it. It is that simple. That is why PostGray can be a 'bit stupid'
with some mail systems that use multiple relays on resends. 

Regards, 

Marco 

On 2020-03-03 14:31, Henrik Morsing wrote:

> On Tue, Mar 03, 2020 at 02:28:55PM +0000, Marco van Beek wrote: 
> 
>> When it connects again it just does the maths and if it is more than 300 seconds it lets it through.
> 
> Since what? If it's stateless, what will it compare the 300 seconds to? 
> 
> It's just impossible, but nevermind, I will figure it out.
> 
> Thanks
> Henrik
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.lug.org.uk/pipermail/gllug/attachments/20200303/94ab1a43/attachment-0001.html>


More information about the GLLUG mailing list