Hello, On Thu, Jan 06, 2005 at 09:15:36AM -0600, Michael Parker wrote: > > I have three (basically) identical servers running spamassassin, each > > with their own Bayesian database. All three have equal MX values, so > > each receive the sameish proportions of ham/spam. > > > > Is there a way of, say daily, merging the databases together? I have not > > seen any documentation on this. > > There is a good reason there is no documentation, see this message for > more details: > http://mail-archives.apache.org/eyebrowse/[EMAIL PROTECTED]&msgNo=20337
Makes sense, thanks! I suppose my other options then, if it is worth it, is to just rely on the database from one machine and copy to the other two machines once a day/month/(choose favourite time period here). Or maybe the three machines can send all spam learn requests to the SA running on one machine? The other, perhaps better, option is that I just run any manual learning on all three machines to update all their databases, and ignore automatically learned differences between them. > > I would guess that something like the output of "sa-learn --dump all" > > from each server could be merged, and then uploaded with "sa-learn > > --restore"? > > FYI, --restore will not read the --dump all format. OK, thanks. It was just a guess! Matthew -- Matthew Newton <[EMAIL PROTECTED]> UNIX Systems Administrator, Network Support Section, Computer Centre, University of Leicester, Leicester LE1 7RH, United Kingdom
