Hey all, wondering if anyone can shed a little light on a problem. Recently I installed Dual pIII 550's into an ASUS p2b-ds board Running Slackware 7.0 with 2.2.16 kernel IDE hardrives AIC7890 SCSI adapter, matrox video blah blah.. Anyway since then I have been having one Kernel Oops about every day. 9 out of 10 times its a problem in kswapd (Reported by ksymoops). At the time of the error the system load is fiarly minor, sometimes it's almost 0.05 or as much as 2.00 Reading through some mailing lists I saw that this sort of this was a problem with a lot of smp systems. (Dating back to the 2.2.10 Kernel) Those messages mentioned something about the problem going away if they rolled back to 2.2.5 On a lark I did a ' swapoff -a' then mkswap -c -v1 /dev/hda7 Didn't get any errors so I checked it again Eventually I did 'swapon -a ' and tried to up the load on my machine..... Well for the next hour I was able to run. * a full tape backup via KBackup * 4 sessions of dnetc at a "nice" level of 5 * 3 xmame sessions 2x in size. * recompiled kernels 2.2.16 ,2.4.0test 5 ,ssh 2.2.0, apache, php, xmms , basically any source packages I had at the time over and over again at the same time Long story short I maintained a CPU load of 25.1 with 53 megs of swapspace being used...... Only thing I noticed was that I had reattched a swap partion with Priority -3 instead of the -1 I normally get at boot. (According to 'dmesg' ) Rebooting and thusly setting the priority back to -1 I was able to crash the machine right away.. I think I got to about 3 on the cpu load and 1200K on the swap partition.... Again I rebooted and got myself back to Priority -3 via the method above and ramped the machine up to a load of around 23.4 Any idea's why the swap priority makes a difference??? -- ----------------------------- |\/|ike at GetBent.net --------------------------------------------------------------------- To unsubscribe, e-mail: tclug-list-unsubscribe at mn-linux.org For additional commands, e-mail: tclug-list-help at mn-linux.org