Every now and then, I get a server load high warning from a user stats run and it's not always the same user. I usually get them in the morning after the event has taken place.
When I log in through shell and run top -c the actual load is usually less than 1.0 but the averages are showing what it was when it was high which was over 50. Everything is back to normal by the time I get the email.
Checking the process and usage, there is nothing running that is using that high of a load anymore so it's like the load averages are stuck and not updating.
Other than rebooting, is there a way to unstick the load averages?