[ale] questions about load averages

Alexander Barton alex at cad.gatech.edu
Tue Nov 18 15:30:51 EST 1997


> 
> On Mon, 17 Nov 1997, Geoffrey Myers wrote:
> 
> > The man page for top say that the load average numbers are: "the average
> > number of processes ready to run during the last 1, 5 and 15 minutes."
> > Question is, does this mean they are sitting in the run queue?  I'm
> > trying to understand what these numbers mean.  What's high?
> 
> [...]  On a single CPU machine a load average of one means
> that the CPU is saturated. [...]

(I may be wrong, but....)
A load average of one doesn't mean the CPU has to be saturated.  You could
have a load average of >1 and still have less than 100% saturation.
In practice, it's pretty hard to get all the way up to 100%.
Running a very CPU-intensive process (with no I/O, no disk waits)
could do it.  If this was the only process running on your system,
it would push the load average up to 1.0.  Running two such processes
would keep saturation at 100%, but push the load average to 2.0.
(Still assuming a single processor system.)

Let's say we start 10 CPU intensive processes which run for exactly 10
seconds, then exit.  After that, the CPU is idle for the next 50 seconds.
The load for the first ten seconds would be 10.0, then 0.0
for the rest of the minute.  Saturation for the first 10 seconds would
be 100%, then drop off to 0% also.  The load average for that minute would
be 1.66 while average saturation would be 16%.

An instantaneous load > 1 means that there are processes waiting for
processor time.

Just for fun, I started four finds, plus top, plus everything else
already on my desktop.  find (over NFS) is very not-CPU-intensive.
I got a load average fluctuating around 1.74 while saturation
was around 35% (100% - idle%).

I interpret this to mean that some of the time file system
information arrived at my machine to be processed by one
of my finds.  Occasionally, this info arrived but had to wait
because one or more finds were already using the CPU.  But, for most
of the time (65% idle), all four finds and top and everything else
was waiting on I/O.

> -----
> Jason Boyles <jason at alltel.net>
> "Those who consider arithmetical means of generating random numbers are,
> of course, in a state of sin." --J.V.N.
> 
> 

-Alexander






More information about the Ale mailing list