We would love to be able to do something like that. Currently we do, like you, per host or per page ad hoc LOAD specs. Typically we watch a system for a few weeks to find the 'normal' load and adjust the config accordingly. If we could have a default that is calculated based on number of cpus/thread, it would go a long way to not needing per host specs.

Steve Holmes Purdue University

On Tue, Jan 31, 2012 at 3:47 AM, Nicolas LIENARD <nicolas at lienard.name>wrote:

Hello

Our LA rule is currently "LOAD 5 10" but we can't apply this kind of rule automatically for all our devices anymore.

Indeed, we have more and more devices with many threads.

I'd like to know if people have the same issue because it's very annoying to override the LOAD when we could automatize it.

For instance we would like to follow this kind of rule:

1 cpu : "LOAD 2 4"

2 cpu : "LOAD 4 8"

2 cpu, 12 threads by cpu (=24 threads in total) : "LOAD 48 96"

We could calculate it with n=grep -ci process /proc/cpuinfo and applying LOAD "n x2" "n x4"

It is possible to have this kind of feature ?

Anybody else is interested by that ?

how do you handle your Load Average threshold ?

CHeers Nico ______________________________**_________________ Xymon mailing list Xymon at xymon.com http://lists.xymon.com/**mailman/listinfo/xymon<http://lists.xymon.com/mailman/listinfo/xymon>

-- If they give you ruled paper, write the other way. -Juan Ramon Jimenez, poet, Nobel Prize in literature (1881-1958)

I prayed for freedom for twenty years, but received no answer until I prayed with my legs. -Frederick Douglass, Former slave, abolitionist, editor, and orator (1817-1895)