Subj : 1:10/345
To   : Matt Bedynek
From : mark lewis
Date : Mon Dec 03 2001 03:26 pm

 MB> I guess its a matter of how well the system is configered.  To
 MB> this date not one of my FreeBSD have ever broke to the level
 MB> it needed to be power cycled in order to recover.  And I can
 MB> assure you that I handle much more load than your typically
 MB> acostomed to.

 ml>> suppose you explain to me how one of my machines, during heavy disk
 ml>> access loads can manage to fire off several dozens of the same task?
 ml>> also explain to me how none of those tasks can be accessed or
 ml>> terminated without shutting down the entire machine...
 ml>> anyfrickingthing is possible when a machine attempts to do
 ml>> toodamnedmuch and runs into a resource problem where resources are
 ml>> controlled from a "central" point in the system...

why is it that it appears that certain keywords are missed and/or interpreted 
at times?

i have disk loads here that i've seen put my CPU usage to 100% for several 
minutes at a time... this causes adverse effects to other tasks, native or not 
is beside the point i'm attempting to make...

i have other tasks that, at times, go into some tight loop chewing up 100% of 
the cpu... if i'm away from the systems, this can last for days ;-( and i know 
that it has knocked my cron program out of sync...

i'm not talking about any winstuff or any nixstuff... it doesn't matter as it 
is beside the point that this type of stuff can and does happen and can wreak 
havoc on systems and the processes they execute...

just to reiterate:

   the point is that this type of stuff can and
   does happen and can wreak havoc on systems
   and the processes they execute...

period, end of statement, end of point, end of thread.

)\/(ark

 
* Origin: (1:3634/12)

.