Distributed bayes

[ian douglas]

> I have 2 servers acting as the gateways for our email domains. I am
> working on syncing the config files between them automatically and would
> like to do the same with bayes, but it seems  little bit beyond my
> skills.

Perhaps copy the mail archives of known spam/ham to each machine and run
local sa-learn processes to update the bayes db's on each system with fresh
spam/ham.

I do this myself and it works great.

-id