On 01/07/2012 10:10 PM, Todd Millecam wrote:
> Unix load is a tricky thing to pin down. It's both a measure of your
> processor power, and of your RAM and how fast the two can talk to each
> other. Also, depending on how things were compiled, the load either
> can mean 1 for all processors being utilized, or have a load of 1
> added on for each processor on the system (so a quad core would have
> either a max load of 1 or 4)
Neither actually. It's the length (possibly aggregated across CPUs) of
the WAIT queues. The maximum load average you could have is the total
number of processes that you are running. I've had NFS server deadlock
issues on Solaris before and the load average climbed to over 300, with
0% CPU utilization (which is logical as if all 300 processes are stuck
waiting, nothing is going to be running!).
Trying to start a Gnome session on Linux with a slow disk (say NFS over
a network with issues) can bring up a load average to 20!
Is a high load average bad? Not necessarily, but normally a high load
average means your CPU is not being utilized at it's maximum capability.
The opinions expressed in this message are the responsibility of their
author. They are not endorsed by BYU, the BYU CS Department or BYU-UUG.
List Info (unsubscribe here): http://uug.byu.edu/mailman/listinfo/uug-list