adverse effect of SA cache

[Matt Hampton]

jh wrote:
> Hi,
> 
> I experienced a bad effect of SA cache:
> As I got hit by constantly more than 500 connections from
> multiple machines trying to deliver the same message,
> no message got rejected even after updating bayes, body rules etc..
> A commandline call to spamassassin rejected all messages,
> but obviously the first of these messages got not identified
> and blocked the cache for all subsequent calls to SA.

Why didn't you just shorten the cache lifetime in the config?

matt