1

I am having an i3 linux Server and i got the below mentioned info about the system

root@XYZ:~# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 42
model name      : Intel(R) Core(TM) i3-2120 CPU @ 3.30GHz
stepping        : 7
microcode       : 0x25
cpu MHz         : 1600.000
cache size      : 3072 KB
physical id     : 0
siblings        : 4
core id         : 0
cpu cores       : 2
apicid          : 0
initial apicid  : 0
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes

The current Load Average is and uptime is: load average: 3.70, 3.03, 2.77

I want to monitor the load average of the server through Nagios. For that i want to know about the optimal load average for the system to set the warning and critical threshold.

i got some information : i7 = 4 cores with 2 threads per core if its like that, then what will be the maximum load average threshold? 8 or 4?

tshepang
  • 12,111
  • 21
  • 91
  • 136
Rakesh
  • 41
  • 1
  • 7

1 Answers1

4

Optimal Load average equals your number of CPU Cores.

if you have 8 CPU Cores (can be found using cat /proc/cpuinfo) on a Linux server, the ideal Load average should be around 8 (+/- 1). If its > 8, then the server resources are over-utilized and if < 8, the server isn’t running with its full potential.

Joyal George
  • 127
  • 10
  • 1
    Currently I am having 2 cores. and my load average is 3.70, 3.03, 2.77. Also my server is working good now. So its not possible to set 2 as the warning/critical threshold in Nagios because it will generate lot of Alerts. – Rakesh May 20 '14 at 07:35
  • What i said is optimal load average is equal to the cpu cores. If your system uses more than 3 the it's over utilizing the resources. If it works fine even at 3.70 then set load average to above 4. – Joyal George May 20 '14 at 07:50
  • you can check your load average using command tload in graphical mode. – Joyal George May 20 '14 at 07:51
  • I have configured as per your comment: "Optimal Load average equals your number of CPU Cores. if you have 8 CPU Cores (can be found using cat /proc/cpuinfo) on a Linux server, the ideal Load average should be around 8 (+/- 1). If its > 8, then the server resources are over-utilized and if < 8, the server isn’t running with its full potential." – Rakesh May 20 '14 at 10:23
  • 2
    This is an over-generalization, and only applies if your workload consists primarily of CPU-bound tasks. If you have a lot of I/O (disk, network, user or anything else), it's entirely reasonable to have a load average that is considerably higher than your CPU/core count, yet have near zero actual CPU utilization. Load average is probably not the metric you want to be watching. – twalberg May 20 '14 at 17:59