Charting server load

[Lennart Sorensen]

On Fri, Feb 02, 2007 at 01:21:30PM +1859, Ian Petersen wrote:
> I believe, and someone can correct me if I'm wrong, that "maximum
> load" is the same as the number of CPUs in the system.  In other
> words, if you have a uniprocessor, and the load is 1, then the CPU is
> always busy, but you haven't overloaded the system.  In a
> dual-processor setup, you can run with a load of 2 and it's the same
> as a uniprocessor running at 1--both processors are maxed, but the
> system's not overloaded.
> 
> So, assuming my understanding of the load number is correct, your web
> servers would need to be 40-processor machines to be able to run at a
> load of 40 without being overloaded.  Or, turning that on its head,
> your hosting company has apparently oversold the server by a factor of
> 10 to 40, depending on how many cores are in the machine, which is
> unreasonable in my eyes, unless its costing you pennies.

Well not quite.  A load of less than 1.00 (on a single cpu system) means
you are not running at 100% cpu load.  On the other hand you could have
the system running at 100% cpu, but having it running 50 tasks at the
same time, which would be a load of 50.

The load is simply: On average how many processes are in the 'ready to
run' state.  If on average, 0.5 processes are ready to run, then you can
only be using 50% cpu, while if 3 processes are, then you must be
running at 100% cpu, unless you have more than 3 processors.

Of course a system with a load of 50 might be doing fine and be
perfectly responsive, while another with a load of 5 might be bogged
down with too much memory and disk io and be very slow at responding.

--
Len Sorensen
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists