[prev in list] [next in list] [prev in thread] [next in thread] 

List:       exim-users
Subject:    Re: [exim] greylisted due fallback server
From:       "John W. Baxter" <jwblist3 () olympus ! net>
Date:       2007-08-27 18:06:16
Message-ID: C2F860A8.14545%jwblist3 () olympus ! net
[Download RAW message or body]

On 8/25/07 2:16 AM, "Leonardo Boselli" <leo@dicea.unifi.it> wrote:

> In sending to a list i noticed my messages were greylisted and delayed a
> lot of time.
> It turned that they used a system that refuse with a 4xx error any message
> the first time that arrives from an "unknown" server and after 5 minutes
> it begin to accept mail fron that server.
> It did not worked for me since in my setup I have:
> 
> begin routers
> 
> lookuphost:
>   driver = dnslookup
>   domains = ! +local_domains
>   pass_on_timeout
>   self = fail
>   fallback_hosts = vettore.xxx.it
>   transport = remote_smtp
>   no_more
> 
> smart_route:
>   driver = manualroute
>   domains = ! +local_domains
>   route_list = * vettore.xxx.it bydns
>   transport = remote_smtp
>   no_more
> 
> --
> 
> this way if for some reason the connection cannot be get on the first time
> the messages are passed to the server vettore, that is on a different
> network so it has different chanches to connect.
> Unfortunately this is setup so vettore know that if a message arrive to it
> it has passed a first try, so when it tries again it know is the second
> try and so the delay between the second and the third try is one hour.
> It happens that some messages bounced with 4xx are then passed to the
> secondary server, so the messages are tried first time by main server,
> after 5 minutes by secondary server, and then each hour, thre hours, etc.
> In this case is possible that the greylisting count is reset so the
> message never passes !
> 
> I contacted the administrator of the remote server and they replied that
> my server does not respect RFC since after a 4xx does not retries but just
> pass the burden to a different server, that in turd has a larger timeout.
> This is not clear to me. Is it true ? I knew that timeouts are minumum,
> not maximum, and that i never found a RFC saying that if a remote server
> says 4xx you con retry only with the same source address.
> Any answer ? If i am wrong, how to change the config file ?
>  
>  
> 

Your system would eventually get messages through our particular greylisting
system.  The first try would time out, as it is never retried.  The second
(overall) try would start a new sequence, and get a temporary failure.  If
the third try (second try by the second server) is from the second server,
same IP, same <mail from:> [after some normalization we do], same recipient,
it would go through.

But greylisting implementations differ--the only way they are viable given
all the odd servers in the world is by providing for whitelisting.

  --John



-- 
## List details at http://lists.exim.org/mailman/listinfo/exim-users 
## Exim details at http://www.exim.org/
## Please use the Wiki with this list - http://wiki.exim.org/
[prev in list] [next in list] [prev in thread] [next in thread]