[NTLUG:Discuss] Top question

[Kelledin]

On Wednesday 21 January 2004 06:49 pm, Greg Edwards wrote:
> I've been trying to find out exactly what the load averages
> displayed by top really mean.  Not much luck finding any info.
>  Just about every other aspect of the top output is described
> in detail, but not the load averages.

Load average (loadavg for short) describes how much of the CPU is 
in use, ranging from 0.0 (fully idle) to 1.0 (fully saturated) 
and above.  It goes above 1.0 if processes marked "ready to run" 
end up waiting until a CPU has some free time.

(FYI, I've gotten loadavg up above 330 during stress-testing.  
The system pretty much became unresponsive after that, so top 
wouldn't give me any further readings until the load died down. 
;-)

If you have multiple CPUs, I believe the loadavg is taken as the 
"average" of the values measured for each processor.

IIRC, the first loadavg number in "top" is the average loadavg 
over the past minute.  The second is the average loadavg over 
the past five minutes.  The second is the average loadavg over 
the past fifteen minutes.  You get the same loadavg measurements 
using the "uptime" command.

-- 
Kelledin
"If a server crashes in a server farm and no one pings it, does 
it still cost four figures to fix?"