Συντάκτης: Stuart Gall Ημερομηνία: Προς: Mark Adams Υ/ο: exim-users, Marten Lehmann Αντικείμενο: Re: [exim] problem of understanding of unseen, self=pass and more
On 18 Oct 2006, at 02:59, Mark Adams wrote:
> Further to this..
>
> On Wed, Oct 18, 2006 at 02:32:10AM +0300, Stuart Gall wrote:
>>
>> On 18 Oct 2006, at 01:59, Marten Lehmann wrote:
>>
>>> Hello,
>>>
>>>> The first router matched will be processed, unless unseen.
>>>
>>> thats exactly the problem: I one router is successful, I have to set
>>> "unseen", otherwise the following routers aren't processed. But at
>>> this
>>> time I don't know if other routers will be successful. So if I'm
>>> setting
>>> "unseen", then exim behaves as if this router has failed altough it
>>> will
>>> process the message with a transport if any following router is
>>> successful.
>>> But if none of the following routers is successful and the first
>>> router
>>> is "unseen", then exim will reply "Unroutable address". Thats the
>>> problem.
>>>
>>
>> How about putting a copy the first router at the end without unseen
>> and configuring a transport to blackhole anything that comes through.
>>
>
> Marten, to have it work correctly you need to have a router to shadow
> all mail, if it copies just the first router then other subsequent
> unseen routers will bounce unrouteable addresses also. a router like
>
> shadow_all:
> debug_print = "R: shadow delivery"
> driver = accept
> transport = shadow_delivery
>
> and the transport (completely untested)
>
> shadow_delivery:
> debug_print = "T: shadow transport"
> driver = appendfile
> file = /dev/null
OR
blackhole:
driver=redirect
allow_blackhole
data = :blackhole:
Presumeably with some domains = specification in which case you may
(or may not) need to set more=no
>
> Of course this will mean no messages will bounce. Correct messages
> will
> be caught by correct routers, everything else will fall into the
> ether..
