> On Jul 7, 2016, at 1:10 PM, Jim Shupert <jshup...@pps-inc.com> wrote: > I am wondering what a "wise" method of doing the catch all account regarding > spam might be > > To limit the amount of spam that a standard user who is catch all (me for > example ) > > I have created a usr named d...@mydom.com <mailto:d...@mydom.com> > this "usr" has a quota of 40 MB … so it goes over quota in a day or so... > It is ,,,,,for the sake of argument ,,,,, ALL spam. > what are you wise folk doing?
Because spammers will spam anything and everything — I have seen spam targeting ‘email addresses’ that were obviously created by some scraper program so dumb that it thought a message ID (something like “122324313109.1231...@somedomain.com”) was an email account — I would question whether there’s any value in having a catch-all. Better to set up .qmail files for the addresses you actually want, and then just send everything else to /dev/null. To do that, create a ‘.qmail-default’ file for your domain, enter a ‘#’ character on the first line, and then add one blank line after it. If you think that you might some day get useful mail sent to a catch-all address, then you’ll probably want to do two things. One is to set up a cron job that just throws away everything in the catch-all account at regular intervals, so that you don’t go over quota and start bouncing mail. The other is to use something like procmail to filter the mail coming into the catch-all. You can write two kinds of filters. One filter will throw away stuff that’s known to be spam (to prevent the mailbox overflowing, and to reduce the amount of mail you need to review manually). The other should look for particular keywords that indicate something that might be interesting to you, and divert that to one of your active mailboxes. Also consider making use of Spamdyke features — for example, rejecting messages from domains without valid RDNS — to reduce the amount of spam you need to process. Angus
