Archive for August, 2011

How the UNIX system Load Avarage is calculated on Mac OS X and Linux

Wednesday, August 24th, 2011

The Load Avarage values that are shown using the ‘top’ commad or the ‘uptime’ command represent the number of blocking processes in the run queue averaged over a certain time period.

An example fo HIGH Load Avarage:

 load average: 12.54, 12.71, 12.19

these values represent CPU, Disk I/O and Network I/O.

if you run the ‘top’ utility you can seek for the CPU usage and for the CPU I/O waiting time that are represented by the ‘us‘ and ‘wa‘ abbriaviations.

If these uage is permanently around 100% then chances are the problem is related to your CPU.

If the I/O waiting time is mostly above the 80% it means that there could be some problem with the I/O devices suche network cards or failing hard drives.

In my spefic case I notices that in ReadyNAS with 6 disks in RAIDX-2 the load avarage is constatly over 10.00,10.00,10.00. I believe this is normal considering that the ‘md’ process has to keep the RAID chain continuosly checksummed and alligned.

This information has been sources from Andy Millar’s really nice blog I have found digging around: http://www.andymillar.co.uk/blog/2006/12/24/linux-load-average-explained/