Use rblsmtpd to tag messages rather than blackholing?
I'm started using rblsmtp to blackhole messages from sites listed in a variety of open-relay and other anti-spam DNS services. In "run": /usr/local/bin/tcpserver -v -u 82 -g 65534 0 smtp \ /usr/local/bin/rblsmtpd \ -r inputs.orbz.org \ -r outputs.orbs.org \ -r or.orbl.org \ -r relays.ordb.org \ -r dev.null.dk \ -r orbs.dorkslayers.com \ -r orbs.gst-group.co.uk \ -r relays.osirusoft.com \ /var/qmail/bin/qmail-smtpd 2>&1 | \ /var/qmail/bin/splogger qmail-smtpd 2 I've noticed some legitimate list mail disappearing and see some notes in the logs about other rejections, but I can't get a sense of what's being rejected. Is there a way to use rblsmtpd, or some other tool, to mark a message as potential spam, along with a message like the one it logs like: rblsmtpd: 24.0.95.144 pid 11121: 451 IP address 24.0.95.144 is an open mail relay or part of a multistage open relay - See http://www.orbl.org If it could instead of /dev/nulling these messages simply add an "X-header" I could have my MUA file them to a "suspicious" mailbox and see what I'm missing -- at least until I get comfortable enough for it to blackhole this stuff, sight unseen. Thanks.
Distributed Checksum Clearinghouse (DCC) antispam for qmail?
With ORBS recent demise and the commercialization of MAPS, I started looking for other antispam measures. The most promising I've found is the Distributed Checksum Clearinghouse: http://www.rhyolite.com/dcc/ If I'm reading it correctly, the code computes a variety of checksums on portions of messages coming through your MTA, and sends these to a DCC server which keeps running counts of each reported checksum; spam sent to a wide audience would increment the same sums so you could detect it. Clients can query this and decide what to do with any incoming message. It has whitelists so that large list mail (e.g. inet-access) would be excluded from spam consideration. Seems to be built for integration with sendmail. Anyone using it now with qmail? I haven't found anything useful searching google for "dcc qmail".
Re: ticketing system?
On Wed, 4 Apr 2001 23:02:32 +0200, Henning Brauer <[EMAIL PROTECTED]> said: Henning> Jay Jarvinen and I started working on an own ticketing Henning> system, it will appear on sourceforge soon. I'll post a small Henning> note when we have a useable version up if others are Henning> interested. It would be a big win if it were integrated with a source code control system like RCS, CVS, or Perforce. In my work I'm using Perforce and Keystone, and I'm really tired of having to enter the same info both places ("fixed bug with patch blash to file foo.c"). It would be cool if the integration allowed me to pop to a ticket, open broken files, fix 'em, check them in, and the ticket system would automagically update its info like bug status and files affected. Perforce has sponsored work to integrate their souce code control system with TeamTrack and Bugzilla. They have an API for integrating other bug-tracking systems. Might be worth looking at. http://www.perforce.com/perforce/products/p4dti.html
Re: VDomains
On Mon, 1 Jan 2001 19:45:52 -0600, Matthew Patterson <[EMAIL PROTECTED]> said: Matthew> I would recommend the vchkpw package at www.inter7.com. It Matthew> lets you manage several virtual domains on one machine by Matthew> using an authentication mechanism that uses the username and Matthew> the domain as opposed to just the username. Ditto. I just set this up at one ISP and really like it. Virtual domains and virtual users, and you keep all the account info out of /etc/passwd. Use it's modified password-checking thing with the qmail POP server to get users access to their mail. Then grab "qmailadmin" web gui interface to manage it, if you like that sort of thing. Then add sqwebmail to give users web access to their mail. Real nice setup.
Re: Mass Mailout Performance Tips
On Thu, 7 Sep 2000 17:47:37 +0200, Peter van Dijk <[EMAIL PROTECTED]> said: Peter> The real trick to high mailinglist performance is only Peter> injecting a message once. qmail is excellent at high-rate Peter> delivery of one message to 20.000 recipients. Peter> It sucks at handling 20.000 separate messages all injected at Peter> the same time. Can you be more specific on the last bit? It can't suck more than sendmail, can it?
Re: "Multi-RCPT vs. Single RCPT delivery" - logic error?
On Fri, 28 Apr 2000 15:02:06 -0500 (EST), Chris Hardie <[EMAIL PROTECTED]> said: Chris> we do web development for an organization that has a PR firm Chris> develop brochures and then send them to us for posting on their Chris> website. The files are often 7-10 MB in size, large enough to Chris> be cumbersome for e-mail, small enough to make overnighting a Chris> ZIP disk seem a little excessive. [...] Chris> I'd be interested to hear if anyone's found a good general Chris> solution to this in a production/business environment. There are a number of sites out there which allow WAN-based file sharing. At the risk of getting flamed... I worked on one which is targetted at exactly this problem. The "sender" uploads the file via WWW, selects who he wants to "send" it to, they get an email notification and download the file via the URL provided in the mail. It ain't rocket-science and folks have found it easy to use. It's totally free, there's currently no advertising, and only the sender has to get an account to upload. Currently user space is 50MB and files expire after 7 days. See: http://www.WhaleMail.com/
Re: methods for ETRN
On Tue, 26 Oct 1999 03:32:57 GMT, Sam <[EMAIL PROTECTED]> said: Sam> If you want reliable mail delivery, use a permanent, reliable Sam> transport, and run SMTP on top of it. Sam> If you have part time connectivity, use any kind of a part time Sam> mail transfer protocol, such as POP3, IMAP, or UUCP. This is impractical for many sites where 24x7 connectivity is prohibitively expensive, and where the organization is UNIX-clueless (e.g. a MicroSoft shop). At such sites, the client organization typically has a machine which dials into the ISP once or twice a day. Queued up mail on the Exchange server gets transfered to the ISP and likewise they want mail which has queued on the ISP mail server to start feeding into their LAN server. Yes, this is very much like the way UUCP is used but I haven't seen MS Visual-UUCP hit the market yet :-) And if it did, the MS Administrator would probably not have the skills to configure it. ETRN is a bit of a hack and a security concern but it does work. I'd love to hear other suggestions for situations like the above where there's not full-time connectivity and they don't have UNIX/UUCP gurus on staff.