On Wed, 18 Nov 1998, Doug S wrote:
> But when I connect with Netscape right afterwards, I get a refusal.
> >From my rejectlog:
>
> 1998-11-18 10:38:19 refused relay (host accept) to <dougsyo@???> from
> <doug@???> H=slip166-72-90-241.oh.us.ibm.net (cc.ysu.edu)
> [166.72.90.241]
It obviously did the DNS lookup OK, since it has the correct name there.
Could this be a permission problem on the file? (Sorry if you mentioned
this - I'm doing so many things at once today my brain isn't recalling
short-term stuff at all.)
> I *did* find something tho: even though I don't have either helo_verify
> or helo_strict_syntax set (verified via -bP), if the helo doesn't match,
> it's flagging it as a "bad relay". I checked this with telnet: if I
> say 'helo cc.ysu.edu' it rejects it, but if I say 'helo slip166...' it
> accepts it. Explicitly setting no_helo_verify and no_helo_strict_syntax
> doesn't help.
You mean it rejects the relaying based on the HELO data? That is most
strange indeed.
> Another data point: When I first implemented this, it worked right,
> and those scripts are dated from when I ran 1.90. I put in 2.02 since
> then.
I've looked through the ChangeLog from 2.03 onwards, and there doesn't
seem to anything relevant.
Well, there is one thing you can do. You can run the listening daemon
with debugging turned on while you do a netscape test:
exim -d9 -bd
This will write a lot of stuff to stderr. You can kill it with ^C since
it doesn't disconnect from the terminal in this case. That might give
some clue as to why the -bh test differs from real life.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
