Problem I'm having it seems is things run just fine. Once this happens it's too late. Can't get in through SSH or Webmin, can't find out what's running and choking things out and the only recourse is to reboot. I'm wondering if I need to be looking at a param that also controls how long a socket stays open or on hold once it's no longer being used because that kinda seems to be what the issue is - eventually to many files are left open and the choke hold ensues - if these files were closing properly doesn't seem it would happen.
I also noticed a sysctl kern.maxfiles resulted in 7048 which is lower than any of the other nix boxes so I set this up to 8000 just to see if it helps any. The previous errors were indeed sendmail but those were from the system trying to email errors to the root account which it could not.
Figure it would also be good to note I am running 225 slots on this box with 512 of RAM which seems sufficent (100 Mb of bandwidth) as I don't think the SC process takes too much memory for each slot and it's ran at max a few times with no issue.
Gee, where's a linux guru when you need one