dying children?

[Jon Dustin]

>>> On 7/29/2008 at 12:03 PM, in message <488F3F37.8020607 at fsl.com>, Steve Freegard
<steve.freegard at fsl.com> wrote:
> Jon Dustin wrote:
>>>>> On 7/29/2008 at 11:17 AM, in message <488F3472.1090904 at fsl.com>, Steve 
> Freegard
>>>
>> 
>> Sure enough, I found a message that was causing the MailScanner children to 
> die. Each of my MailScanner hosts had a copy (guess I'm just lucky) delivered 
> on July 7, and that is exactly when performance began degrading.
>> 
>> Julian - Would you like a copy of these messages? They look like spam to me, 
> but I am not sure exactly where the problem with MailScanner lies.
>> 
> 
> What output do you get from --debug?  As it might not be a problem with 
> MailScanner but one of the associated utilities.

Here is the output from --debug --debug-sa

13:27:38 [31635] dbg: learn: auto-learn? ham=0.1, spam=12, body-points=0.288, head-points=0.288, learned-points=-2.599
13:27:38 [31635] dbg: learn: auto-learn? no: inside auto-learn thresholds, not considered ham or spam
Segmentation fault

Anything else to try?

PS - My MailScanner(s) are working much better since I have removed this nasty message.