The host load average seems to be a sum of the VPS laverage:
[root@xx ~]# vzlist -o laverage,cpulimit
LAVERAGE CPULIM
0.00/0.00/0.00 0
0.00/0.00/0.00 0
3.54/2.82/1.79 26
0.00/0.00/0.00 0
0.00/0.00/0.00 0
top - 05:12:05 up 15 days, 19:49, 1 user, load average: 3.56, 2.82, 1.78
Tasks: 291 total, 3 running, 288 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.2%us, 6.3%sy, 0.0%ni, 92.9%id, 0.1%wa, 0.0%hi, 0.5%si, 0.0%st
Mem: 8140924k total, 6908692k used, 1232232k free, 647508k buffers
Swap: 2031608k total, 28k used, 2031580k free, 5522328k cached
When we limit the cpu the host load average gets higher
top - 05:12:05 up 15 days, 19:49, 1 user, load average: 3.56, 2.82, 1.78
but the real usage get's lower (as it should)
Cpu(s): 0.2%us, 6.3%sy, 0.0%ni, 92.9%id, 0.1%wa, 0.0%hi, 0.5%si, 0.0%st
Shouldn't the host load average be calculated based on the host stats instead of the vps stats?