Good. :) And now that Sendmail::Milter is out of the loop, I doubt you
will ever see it crash. I have been running SRS-socketmapd (based on more
or less the same programming frame) for over two months now; and it has
never crashed so far, either.
I really appreciate you testing the waters; and I would really like to
hear back from people like you, in case SPF-socketmapd behaves
unexpectedly (which is doubtful too, but still).
I've used the SRS stuff in the past and it works fine.. Might enable that as
well later..
A few problems tho:
1) no easy way to easily put into sendmail.mc! (had to change a few things
around in the sendmail.cf after the m4) (maybe the reason for 2)
2) in the log file I am getting a lot of:
Feb 23 20:24:36 dbmail-mx4 sm-mta[11848]: j1N7ORo1011848:
ruleset=check_mail, arg1=<noreply(_at_)mark(_dot_)tanea>,
relay=dialup-38.69.221.203.acc07-wick-bne.comindico.com.au [203.221.69.38],
reject=550 5.7.1 <noreply(_at_)mark(_dot_)tanea>... Please see
http://spf.pobox.com/why(_dot_)html?sender=noreply(_at_)mark(_dot_)tanea&ip=203.221.69.38&rec
eiver=dbmail-mx4.orcon.co.nz:domain of sender noreply(_at_)mark(_dot_)tanea
does not
exist
Its not a SPF problem.. really.. the domain they are sending from does not
exist.. but the SPF wants to stick its ugly head in the way and do the tests
for me (and the reason why it was rejected was nothing to do with SPF)
So its doing the SPF stuff and then doing:
R<PERM> $* $#error $@ 5.1.8 $: "553 Domain of sender address "
$&f " does not exist"
Unless I have the SPF check_mail in the wrong place :-( (which I may have)
Anyway NOT to do the "domain does not exist" test in the Socketmap and allow
sendmail to do its normal stuff?