[clug-talk] understanding top output

[Kevin Anderson]

Those load averages should be taken with a grain of salt.  Waiting on IO will 
jack them through the roof, even though the processor isn't doing anything.  
( man iostat for help with IO bottlenecks)

Also a load average of 1 means a 100 percent load on 1 processor.  So if you 
have a single processor machine, and the load average is >75, I'd say it's 
pretty high.  On the other hand, if you have a load average of 5.99 on an 
8way system, you've still got plenty of spare cycles.

Lastly, Linux is MUCH more stable than some other OSes.  Even on a single 
processor system, a load average of 2 or higher (I've seen 100+) doesn't mean 
that the server will crash.  It won't.

Kev.

On Thursday 18 November 2004 15:44, smagnuson at forzani.com wrote:
> > Does anyone here know where to find a detailed description of the "load
> > average" field in top?  What is used to determine these numbers?  This is
> > all I have found so far:
> >
> > ...system load averages for the past 1, 5, and 15  minutes
> >
> > IE:
> > :top
> >
> > top - 14:38:06 up 1 day,  5:35,  5 users,  load average: 0.00, 0.00, 0.00
> > ...
> > ...
> >
> > :top
> >
> >  15:40:18  up 60 days,  9:59,  6 users,  load average: 10.97, 10.44,
> > 10.67 ...
> > ...
> >
> >
> > you can also see it in
> > w
> > uptime
> >
> > thanks
> > sig
>
> _______________________________________________
> clug-talk mailing list
> clug-talk at clug.ca
> http://clug.ca/mailman/listinfo/clug-talk_clug.ca
> Mailing List Guidelines (http://clug.ca/ml_guidelines.php)
> **Please remove these lines when replying