[Mimedefang] Puttting MIMEDefang spool on tmpfs

[James Ebright]

I would take a close look at your log files for the time periods in 
question.... perhaps you have some spambot zombies or virus-remailers 
inside your network or using your box off hours.

lsof and top and ntop will give you alot of useful info also during the 
timeperiod in question.. see whats eating memory, why, etc in case its 
not mail/spam/virus related (although it sounds like spambot runs to me 
anecdotaly, esp if nothing had changed).

Jim

Lisa Casey wrote:

>I think you've hit the crux of the matter here. You might remember that I
>also mentioned that this box with it's Sendmail/Mimedefang has been working
>fine up until weekend before last. That's when I started getting the error
>messages in my log re mimedefang having no free slaves and when that
>happened load averages shot up and Sendmail began refusing connections due
>to the load average. When I brought mimedefang back up today, load averages
>climbed back into the 30's within an hour and the box became sluggish almost
>to the point of not being able to loog into it. It ran perfectly all weekend
>(load averages consistently under 1.0) all weekend  with no mimedefang (but
>lots of customer complaints about increased spam!).
>
>Figuring out WHY mimedefang is suddenly shooting up the load averages when
>it worked so well for quite some time is what I ought to do to slove the
>problem. Knowing why will tell what the cause is.
>
>I'm at home now but tomorrow I'll look at cron jobs, etc. What else should I
>take a look at? There's nothing that I can  think of right off hand that I
>have changed or added around that same time.
>
>I got a quote on Canit today, unfortunately my company can't afford to go
>that route just now  :-(  So Mimedefang/Spamasassin is going to have to be
>it for me for now.
>
>  
>