Re: [Xen-users] XCP, creating a metric and looking for input

[Thomas Halinka <lists@xxxxxxxxx>]

Hi Verne,

Am Samstag, den 10.04.2010, 20:56 -0400 schrieb Vern Burke:
> Greetings all:
>     I'm sitting here on a Saturday night (life in the fast lane) working 
> on putting together good blended host and virtual machine metrics for 
> the upcoming load balancer so I thought I'd ask for some input from the 
> crowd :).

;-)

> 
> My experience is that, at least on my cloud, I'm far more likely to 
> overload memory than CPU. In fact, as far as I can tell, I've never been 
> able to even significantly load the CPUs on any of my (admittedly older) 
> servers (2 are 2x Opteron, 2 are 2x dual core Opteron), even with a 
> pretty hefty load of VMs running.

So you ever had supreme software :-)

> 
> My question is, for those of you running as a real cloud (no funky 
> specialized hardware passthroughs in the server, etc), have you ever 
> been able to significantly load or overload your cpus before running out 
> of memory?

Sure - every Software brings potential risk with it. Coding-Errors could
lead to such a behaviour.

>  The reason for this question is that I'm establishing 
> weighting for the relative importance of the components of the metric.

Sounds interesting. Is there any docs or whatever about your goals,
techniques and thoughts?

there a nice work: http://contentdm.lib.byu.edu/ETD/image/etd2755.pdf

some useful infos, can be found here, too: http://lbvm.sourceforge.net/

> 
> TIA!
> 
> Vern
> 

cu,

thomas


> _______________________________________________
> Xen-users mailing list
> Xen-users@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-users



_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users