Apache SpamAssassin sa-update channels for SARE rules
The SpamAssassin Rules Emporium (SARE) has a bunch of rules for SpamAssassin. Many of them are good rules. Some of them get updated regularly. To update these rules Chris Thielen wrote, and continues to update, a bash script called rules_du_jour that automatically downloads and when necessary updates a server’s SARE rules.
The rules_du_jour script works fine and has been in use by thousands of people for quite some time. One problem, though, is that it only runs on *nix systems. It also needs to be updated when new rulesets are released. Not a significant task by any stretch, but it does add to a mail system’s administrative load.
Apache SpamAssassin 3.1.1 and beyond includes sa-update, a cross-platform Perl program that manages updating rulesets available via “channels” in a light weight fashion using the DNS to track ruleset versions and HTTP to download updated rulesets when necessary. SHA1 sums and GPG are also used to verify downloaded rulesets.
So far the only sa-update channel available is the default channel published by the Apache SpamAssassin Project (updates.spamassassin.org). In early July I created my own channels containing up to date SARE rulesets to gain experience with sa-update and to automate rule updates on my own systems. For some reason I have never used rules_du_jour myself. I think it’s size and need for updating when new rulesets released turned me off of it.
After much discussion about “rules_du_jour vs. sa-update” in early August on the SpamAssassin users’ list in this and this thread, I decided to make my SARE sa-update channels public. The discussion was largely a debate between long time users of rules_du_jour who were discounting the utility of sa-update and the SpamAssassin development team, along with a number of Windows users who can’t use rules_du_jour. I figured making the channels public would attract enough users to either quash fears about sa-update or identify any problems with sa-update. I’ll be happy with either result.
So far there hasn’t been any problems reported. There’s only been a couple people that have reported their satisfaction though. Numbers often tell a better story. In the last week there have been a few hundred diverse IPs that have download files from one or more channels (each SARE ruleset is in its own channel). Browsing the server logs it appears that a number of noteworthy engineering and computer networking organizations have been using the channels on their mail clusters too. Since no one has complained yet, I’m assuming all is well. I know, myself, I haven’t had any problems in the last month or so.
For those interested in using these sa-update channels, a brief how-to is available here. I was planning on writing some more detailed documentation, but it really is as simple (on *nix systems at least) as stated in the how-to. If anyone wants to contribute documentation for use on Windows or other systems that the how-to doesn’t seem to cover it would be welcomed.
Add comment August 19th, 2006
