Folks,
We have our applications running on a redhat 5.8 guest OS on ESXi 5.0
The load average reported by linux is around 5 times what is seen with the same OS and application running on a rackmount server (Not as a VM).
Looking through the VM top and sar outputs, nothing seems to suggest that the VM is overloaded.
The application is also not showing any signs of having any trouble.
The CPU Idle% is around 30-40% and the runq-sz is comparable to the physical server output.
The Host statistics also suggests that the CPU utilization isnt higher than 50-60%.
Is there something missing or wrong in how redhat is calculating the load average?
ESXi Host:
UCS B-Series Blade with dual 6-core CPUs (Hyperthreading enabled) and 48GB RAM
VM:
4 vCPUs and 12GB of RAM
[root@001 /]# sar -q # load average
Linux 2.6.18-308.el5PAE (001) 02/28/2013
12:00:02 AM runq-sz plist-sz ldavg-1 ldavg-5 ldavg-15
12:10:01 AM 5 3308 26.08 23.65 23.29
12:20:06 AM 1 3293 28.37 25.44 24.17
12:30:02 AM 6 3170 23.56 22.38 23.45
12:40:02 AM 19 3239 24.74 22.43 22.91
12:50:02 AM 6 3283 31.29 29.45 25.40
01:00:02 AM 14 3340 26.60 26.47 25.39
01:10:01 AM 36 3246 21.82 23.52 24.44
01:20:01 AM 3 3234 19.70 20.90 22.42
01:30:03 AM 9 3220 21.52 21.18 21.85
01:40:03 AM 0 3178 23.43 25.39 24.01
01:50:02 AM 0 3211 25.82 26.73 25.48
02:00:02 AM 0 3231 18.72 20.24 22.70
02:10:03 AM 61 3161 29.64 25.43 23.80
02:20:02 AM 40 3200 18.77 21.95 22.56
Physical Server:
Quad Core CPU (hyperthreading enabled), 4GB RAM
qa41:~ #top
top - 14:48:42 up 47 days, 3:34, 5 users, load average: 3.28, 4.66, 4.71
Tasks: 759 total, 1 running, 758 sleeping, 0 stopped, 0 zombie
Cpu0 : 0.5%us, 8.9%sy, 7.4%ni, 82.8%id, 0.5%wa, 0.0%hi, 0.0%si, 0.0%st
Cpu1 : 0.5%us, 9.4%sy, 3.9%ni, 86.2%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Cpu2 : 0.0%us, 5.9%sy, 4.4%ni, 89.7%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Cpu3 : 0.0%us, 21.1%sy, 19.1%ni, 55.9%id, 0.0%wa, 0.0%hi, 3.9%si, 0.0%st
Cpu4 : 0.0%us, 20.6%sy, 19.6%ni, 54.4%id, 0.5%wa, 0.5%hi, 4.4%si, 0.0%st
Cpu5 : 0.0%us, 21.6%sy, 22.5%ni, 51.5%id, 0.0%wa, 0.0%hi, 4.4%si, 0.0%st
Cpu6 : 0.0%us, 4.9%sy, 4.9%ni, 90.2%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Cpu7 : 0.0%us, 18.7%sy, 17.2%ni, 60.6%id, 0.0%wa, 0.0%hi, 3.4%si, 0.0%st
Mem: 4138708k total, 3857800k used, 280908k free, 932k buffers
Swap: 7999992k total, 656k used, 7999336k free, 1488236k cached
qa41:~ #sar -q # load average
Linux 2.6.18_308.cisco_PAE.el5 (qa41) 02/28/2013
12:00:02 AM runq-sz plist-sz ldavg-1 ldavg-5 ldavg-15
12:10:01 AM 4 2998 4.40 4.69 4.99
12:20:01 AM 1 3004 3.52 4.75 5.06
12:30:04 AM 2 3003 3.62 4.37 4.82
12:40:05 AM 7 3000 5.96 5.93 5.35
12:50:02 AM 15 3005 9.08 6.19 5.51
01:00:01 AM 86 3000 3.65 4.21 4.76
01:10:01 AM 3 3002 3.47 4.07 4.48
01:20:02 AM 10 3003 6.64 5.20 4.73
01:30:02 AM 2 3005 4.98 5.18 4.91
01:40:02 AM 2 3002 4.08 5.05 5.01
01:50:05 AM 1 3004 6.44 6.55 5.93
02:00:03 AM 1 3007 4.49 4.92 5.49
02:10:02 AM 5 3002 3.79 4.81 5.30
02:20:03 AM 0 3003 4.00 5.07 5.26
02:30:01 AM 9 2998 5.24 6.01 5.72
02:40:01 AM 4 3003 4.99 5.76 5.69
02:50:02 AM 17 3003 6.53 5.75 5.54
03:00:03 AM 0 3003 3.84 4.72 5.19
03:10:01 AM 2 3003 6.20 6.02 5.55
03:20:01 AM 21 3004 4.28 4.77 5.13