Hello,
when it occurs bload has mostly already finished it's work but is still loaded. We have not tried it without bload. The Bayesian is enabled and there are 17 rules.
After looking into 'ks_log.txt' for the entries which were made at a time of crash, we have noticed that 2 mails came at once and their entries are nested. One of the
entries contains 'Read pointer size to'. Here is an excerpt from 'ks_log.txt' (the crash was at 13.02.2007 07:00:36):
…
13.02.2007 00:00:26 - Form: Memo
*SUBJECT:do eleven*<br/>
*FROM:charge is <cgcbass@veng.com>*<br/>
*MAIL_FROM:cgcbass@veng.com*<br/>
*BODY LENGTH: 7251* <br/>
*ATTACHMENT: alliance.gif*<br/>
*Scanning message*<br/>
*P:0,0018*<br/>
Number of rules: 17
Reloading configuration…
freeing rule list
freed rule list
loading rule list…
Number of rules: 17
13.02.2007 07:00:36 - Form: Memo
*SUBJECT:Citibank Deutschland: Mon, 12 Feb. 2007 12:07:01*<br/>
*FROM:"Citibank Deutschland" <sicherheit@citibank.de>*<br/>
*MAIL_FROM:info@awilson.co.uk*<br/>
13.02.2007 07:00:36 - Form: Memo
*SUBJECT:Citibank Deutschland: Mon, 12 Feb. 2007 12:07:01*<br/>
*FROM:Citibank Deutschland <sicherheit@citibank.de>*<br/>
*MAIL_FROM:sicherheit@citibank.de*<br/>
*BODY LENGTH: 5174* <br/>
*Scanning message*<br/>
Read pointer size to BODY LENGTH: 5103
*Scanning message*<br/>
Number of rules: 17
13.02.2007 09:40:58 - Form: Memo
*SUBJECT:it explanation be vocalic*<br/>
*FROM:"Kathleen Short" <zdpproud@everymail.net>*<br/>
*MAIL_FROM:zdpproud@everymail.net*<br/>
*BODY LENGTH: 7844* <br/>
*ATTACHMENT: malleable.gif*<br/>
*Scanning message*<br/>
Read pointer size to P:1,0000
*Copying to goodmail*<br/>
*ALLOW*<br/>
13.02.2007 09:41:05 - Form: Memo
*SUBJECT:Or casket be sally*<br/>
*FROM:"Desiree Mcclure" <rvdevilish@chenbiao.com>*<br/>
*MAIL_FROM:rvdevilish@chenbiao.com*<br/>
*BODY LENGTH: 7809* <br/>
*ATTACHMENT: catv.gif*<br/>
*Scanning message*<br/>
*P:1,0000*<br/>
*Copying to goodmail*<br/>
*ALLOW*<br/>
…
There are nested entries at a time where no crash happened, but these do not contain 'Read pointer size to'. So, we assume it could be involved in the problem. Maybe, the problem occurs in 'ks_load_bayesian_data()'?
Best regards