smmsp is the sendmail submission queue runner.  It's a seperate sendmail
process that runs /var/spool/clientmqueue, which I believe is where
Sendmail drops outbound messages before delivery (for security through
privelege seperation).  You might have a bunch of messages in the queue
that it's trying to process.

If that queue is empty, use lsof or fstat to see what files the smmsp user
has open when the problem starts.

On Thu, 4 Sep 2003, Jim Streit wrote:

> I have recently noticed a weird thing happening on a server that I have. 
> A few times a day, the drives on my server start thrashing, like they are
> doing a massive read or write.  If I look at TOP while this is happening
> the only thing that really shows active CPU cycles is sendmail.  It uses
> about 8% of the processing power of the box.  After about 10 minutes, the
> thrashing stops and that sendmail thread stops (user smmsp).  During the
> thrashing there isn't any noticeable performance slowdown.
> 
> The Box...
> IBM Netserver 5000
> 2 GB RAM
> Dual PIII 550
> 3x18 gb hot-swap scsi drives (RAID 5)
> Redhat 9.0 (with all of the latest RedHat up2date updates)
> Sendmail 8.12.8-6.90
> 
> Anyone have any ideas on what my be causing this?
> 
> --
> Jim Streit
> Partner & Co-Founder
> NorthLANs Alliance, LLC
> JimStreit at northlans.com
> 
> 
> 
> 
> _______________________________________________
> TCLUG Mailing List - Minneapolis/St. Paul, Minnesota
> http://www.mn-linux.org tclug-list at mn-linux.org
> https://mailman.real-time.com/mailman/listinfo/tclug-list
> 

Adam Maloney
Systems Administrator
Sihope Communications


_______________________________________________
TCLUG Mailing List - Minneapolis/St. Paul, Minnesota
http://www.mn-linux.org tclug-list at mn-linux.org
https://mailman.real-time.com/mailman/listinfo/tclug-list