Re: Changes to /etc/mail/spamassassin/sa-rules.cf not taking after restarting spamassassin service

As best as I know, I invoke spamassassin via the mimedefang hook.



> On Jan 7, 2026, at 8:14 PM, David B Funk <[email protected]> wrote:
> 
> Philip,
> 
> There's probably some other mechanism that is using SA in some way which you are not 'looking' at.
> For example, AMAVISD imports the SA perl "engine" within itself and doesn't use 'spamd' at all.
> You may have some kind of filter-chain in your MSA (postfix maybe?) that is invoking "spamassassin" as a filter process and configured to use SA from some other directory.
> 
> Try stopping your 'spamd' and send yourself mail. If it's still getting filtered then you haven't found the actual SA agent, time to do some detective work.
> 
> 
> On Wed, 7 Jan 2026, Philip Prindeville via users wrote:
> 
>> Hi all,
>> 
>> I'm on Fedora 43 running spamassassin-4.0.1-7.
>> 
>> A rule involving sendgrid.net was causing false positives so I tried to comment it on in spamassassin/sa-rules.cf and restarted the service.
>> 
>> No dice.  I was still getting matches against it.
>> 
>> So I changed the name and lowered the score and did a service stop/start just in case.  No change!!!  The rule by the old name was still getting matches.
>> 
>> What's going on?  That's not supposed to be able to happen.
>> 
>> What would cause it to hold onto an old version of the file even after restarting?  Is there a cache I'm not aware of?
>> 
>> I did a "ps -ef | grep -e spamd -e spamassassin" after the "stop" just to make sure the processes were all being cleaned up.  They were.
>> 
>> So how is this happening?
>> 
>> Thanks,
>> 
>> -Philip
>> 
>> 
> 
> -- 
> Dave Funk                               University of Iowa
> <dbfunk (at) engineering.uiowa.edu>     College of Engineering
> 319/335-5751   FAX: 319/384-0549        1256 Seamans Center, 103 S Capitol St.
> Sys_admin/Postmaster/cell_admin         Iowa City, IA 52242-1527
> #include <std_disclaimer.h>
> Better is not better, 'standard' is better. B{