Autor: wayne Data: A: exim-users Assumpte: [Exim] Re: large SMTP banners and callouts appear to have problems.
In <1072064323.5069.7.camel@???> Kjetil Torgrim Homme <kjetilho@???> writes:
> On Mon, 2003-12-22 at 00:14, wayne wrote:
>> Looking in verify.c do_callout(), there appears to be a 1k
>> responsebuffer and if the banner is larger than this, exim gives up.
>
> I agree with your findings. I note that transports/smtp.c, which is the
> code used for sending e-mail, has a buffer of 4096 bytes.
Is there anything else I need to do to make sure this doesn't get lost
in the mailing list?
For what it is worth, this bug seems to cause exim to report a 4xx
error back, so the sender just keeps retrying until it gives up. I
checked my mail logs going back 3 months and I found two cases where
there were these kinds of repeated retrying due to sender callout
failing. In both of the other cases, telnet'ing into the SMTP port
yields something that just echos back what you send it, not an MTA. I
can't say for sure that this was the case when the callbacks failed,
but it doesn't appear that large SMTP banners are that common.
> I suggest you change the callout code so that they are in sync.
Well, if I actually cared at all about receiving email from the
particular host that triggered this bug, I might be motivated to fix
it in the code I'm running, but I'm not. Long winded SMTP banners
appear to correlate with people who don't have much interesting to
say.
