Re: [Exim] SMTP data timeout

Página superior
Eliminar este mensaje
Responder a este mensaje
Autor: Phil Pennock
Fecha:  
A: Jim Pazarena
Cc: exim-users
Asunto: Re: [Exim] SMTP data timeout
On 2000-09-12 at 12:13 -0700, Jim Pazarena gifted us with:
> I do not have "data_timeout" _or_ "final_timeout" set in my configure
> file, so I therefore must have the 5m and 10m defaults (respectively).
>
> Upon examining my main.log, I see 12% ratio of messages which have
> timed out with:
> SMTP data timeout (message abandoned) on connection from (xxxx.xxx.xxx) [ etc
>
> This doesn't seem right. Should I increase the timeout period, or can
> anyone offer what other problem I may have?


-----------------------------< cut here >-------------------------------
From: Phil Pennock <phil.pennock@???>
Subject: Re: [Exim] Mail problem PLEASE HELP.
To: Exim Users <exim-users@???>
Date: Mon, 4 Sep 2000 20:20:17 +0200
Message-ID: <20000904202017.B28873@???>

On 2000-09-04 at 09:25 +1000, ridgey gifted us with:
> > As far as outgoing is concerned, it appears that any email of size causes
> > the email to timeout and and error returned. Also if the above problem gets
> > to big, it causes connections to be refused that port and no incoming or
> > outgoing to mail be be processed until it is restarted.


> Any help / ideas appreciated.


Firewall filtering fragmented packets, without reassembling them, and
you're also blocking Path MTU discovery somewhere. Likely the same
firewall.

If you don't have access / authorisation to fix the firewall, the only
thing you can do is to lower the interface's mtu (576 should work for
anywhere on the Internet. No really). This sucks.

But some people like firewalls, and don't care if they're broken. :^(
[ snip sig and list waffle ]
-----------------------------< cut here >-------------------------------

--
A science is said to be useful if its development tends to accentuate the
existing inequalities in the distribution of wealth, or more directly promotes
the destruction of human life - Godfrey Hardy, A Mathematician's Apology, 1941