[Cialug] stuck load average and ntp?

[David Champion]

Have a server running CentOS where the date was messed up... yesterday 
the clock was 3 hours slow, I reset it. This morning it thought it was 
August 19 of 2009. I've never been a fan of how RH / CentOS uses ntp... 
instead of running the ntpd service like any other Linux distro, it has 
a cron job that runs "ntpd -q" hourly.

Because of the date issues, it got several cron task queued up, and I 
think that cause the load avg to jump up. I've reset the date and killed 
the cron jobs, but the load avg is still goofy. Also I know the server 
hasn't been up anywhere near 421 days, since it was physically moved 
within that time frame.

Here's what top says:

top - 09:50:54 up 421 days, 12:39,  1 user,  load average: 6.57, 4.42, 3.11
Tasks:  96 total,   6 running,  90 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.2%us,  0.2%sy,  0.0%ni, 77.4%id, 21.9%wa,  0.2%hi,  0.0%si,  
0.0%st
Mem:   1027160k total,   941432k used,    85728k free,    33304k buffers
Swap:  1052216k total,    12216k used,  1040000k free,   663016k cached

The load avg has said exactly the 3 numbers above all morning.

The /var/log/wtmp file had a date of 20090819 too... I did a "touch" on 
it but that doesn't seem to reset anything, did a "true > /var/log/wtmp" 
and that didn't do anything.

Is there another method of resetting the load average, or does it just 
need a reboot?

-dc