* Sloath <[EMAIL PROTECTED]> [2005:11:14:18:44:49+0100] scribed:
> Michael D Schleif wrote on 13/11/2005 07:57:
> >* Tom Collins <[EMAIL PROTECTED]> [2005:10:27:22:30:26-0700] scribed:
> >>On Oct 27, 2005, at 5:02 PM, Michael D Schleif wrote:
> >>>We have a client that runs Groupwise on a Netware server on their LAN.
> >>>We manage everything on this LAN; and, we handle their DNS for their
> >>>domains.
> >>>
> >>>We want to scrub all of their _incoming_ email on our remote toaster,
> >>>doing what it does best, then forward *ALL* of their email onto that
> >>>Groupwise box.
> >>>     
> >>Use simscan.  Make your server the only MX for the domain.  Put the 
> >>domain in /var/qmail/control/rcpthosts (or morercpthosts and run 
> >>qmail-newmrh).  Do *not* add their domain to locals or virtualdomains.  
> >>Add an entry in /var/qmail/control/smtproutes to route the scanned mail 
> >>to their Groupwise server.
> >>--
> >>Tom Collins  -  [EMAIL PROTECTED]
> >>QmailAdmin: http://qmailadmin.sf.net/  Vpopmail: http://vpopmail.sf.net/
> >>You don't need a laptop to troubleshoot high-speed Internet: 
> >>sniffter.com
> >
> >I finally got back on this one.  Yes, setup was simple; and, yes,
> >messages _do_ go through our toaster, and they _do_ end up in the
> >Groupwise server.
> >
> >This setup does *not* appear to work.  I am still waiting for the
> >Groupwise admin to send to me the complete headers for the test
> >messages; but, I am told that *none* of these test messages include
> >spamassassin headers.
> >
> >In fact, the following does *not* show _any_ activity for test messages:
> >
> >   tail -f /var/log/{cl,sp}amd/current
> >
> >I am not clear which process comes first,
> >
> >   /var/qmail/control/smtproutes
> >
> >or simscan ???
> >
> >What do you think?
> >
> The diagram of the found at http://www.nrg4u.com/ does a very good job 
> of explaining the qmail process....
> man qmail-control will also tell what each control file is used by.
> 
> In your case:
> qmail-smtp check incoming email envelope recipient addresses against the 
> contents of rcpthosts to see if it should be accepted.
> It then passes the email onto qmail-queue, unless $QMAILQUEUE is set in 
> which case it will invoke the program named.
> A simple  /home/vpopmail/etc/tcp.smtp should include
> :allow,QMAILQUEUE="/var/qmail/bin/simscan"
> 
> The message is then passed to qmail-send which determines whether the 
> message is local or remote by the presence( or lack of ) the
> envelope recipient in the virtualdomains or locals file.
> If it is determined to be remote, it is passed to qmail-rspawn and then 
> qmail-remote.
> Qmail-remote checks smtproutes for the domain and sends the message as 
> appropriate.
> 
> The short answer was, simscan comes first.
> You turn on debug in simscan by putting the following in 
> /service/qmail-smtpd/run before the call to exec.
> SIMSCAN_DEBUG=2
> export SIMSCAN_DEBUG
> 
> The other thing to check is the /var/qmail/control/simcontrol file
> Make sure clam and spam are set to yes, otherwise they won't be invoked.
> 
> Regards
> Tom Beardshall

OK.

    # cat /home/vpopmail/etc/tcp.smtp
    127.:allow,RELAYCLIENT=""
    :allow,QMAILQUEUE="/var/qmail/bin/simscan"
    24.15.252.147:allow,RELAYCLIENT=""

Remember, this is a functioning toaster.  It has a couple dozen virtual
sites, and hundreds of users.  *ALL* of the functionality, including
spamassassin _and_ clamav, work successfully, and as expected, for these
users.

In fact, doing this:

    tail -f /var/log/{cl,sp}amd/current

I can watch messages pass through simscan, clamav and spamassassin --
provided that they are destined for a virtual user.

Point of fact, GTUBE tests to virtual users result in spamassassin
counts of 999 -- for virtual users; but, I do not see them pass through
simscan; nor are they marked by spamassassin at the Groupwise
destination.

However, the Groupwise administrator cannot seem to get complete
messages and headers to me for inspection.  Does anybody here know how I
can have him extract a message from Groupwise?

What am I missing?

How can I test this further?

What do you think?

-- 
Best Regards,

mds
mds resource
877.596.8237
-
Dare to fix things before they break . . .
-
Our capacity for understanding is inversely proportional to how much
we think we know.  The more I know, the more I know I don't know . . .
--

Attachment: signature.asc
Description: Digital signature

Reply via email to