Load average vs. real CPU load on a multi-core system
807557Dec 10 2008 — edited Dec 12 2008Hi all.
I'm trying to monitor the CPU usage on several systems, and I have run into a problem with the load average. According to the man on uptime, it reports "the average number of jobs in the run queue over the last 1, 5 and 15 minutes". But I'm not sure what that means on our systems.
Could someone tell me how does the classic CPU load average reported by 'uptime' relate to the real CPU load on a system like a T2000, with multiple cores and several virtual CPUs per core?
Let's say we have 8 cores and 32 CPUs. Does a load of, say, 32, 20, 20 mean that each of the CPUs is busy but there are no waiting processes? When could I assume that the machine is CPU stressed?
Thanks in advance for your help.