Re: [Xen-users] Re: [Xen-devel] SLA Monitoring of XEN DomU's

[Tim Post <echo@xxxxxxxxxxxx>]

Sorry for top posting,

At the time I wrote xguests, the sysfs path to determine VBD read /
write / out of requests was in the air. As such, I did not include it.
Adding that is now trivial and an exercise for the reader.

Use it as a (hurried) example to hack the tools you need into working.

Cheers,
--Tim

On Mon, 2010-01-04 at 15:33 +0000, Ahmad Hassan wrote:
> Yeah you are absolutely rite. Do you know any means to record the CPU
> consumption and IO consumption of DomU's on Dom0?
> 
> Thanks for the help.
> 
> Best Regards, Hassan
> http://cern.ch/ahmadh/portfolio
> 
> Erasmus Student
> The University of Reading, UK
> Aristotle University of Thessaloniki, Greece
> The University Of Carlos III, Madrid Spain
> 
> 
> 
> 2010/1/4 Tim Post <echo@xxxxxxxxxxxx>
>         On Mon, 2010-01-04 at 15:11 +0000, Ahmad Hassan wrote:
>         > Thanks for the replies. So I can use either use SNMP
>         monitoring using
>         > CACTI or http://www.pmacct.net for Network traffic
>         monitoring of
>         > DomU's.
>         >
>         > What about CPU availability guarantees for DomU's?
>         >
>         > By SLA guarantees, I mean to say that Whether I can proof
>         and show to
>         > the owner of DomU's that you are getting the exact bandwidth
>         and CPU
>         > you asked for.
>         
>         
>         To do that, give them root access to dom-0 and let them see
>         their
>         running configuration.
>         
>         Beyond that, there is no proof that can't be tainted. Welcome
>         to
>         hosting :)
>         
>         How would you give someone proof that they are one of only
>         five accounts
>         on a shared server?
>         
>         You might want to encourage clients to monitor iowait while
>         doing brief
>         disk operations (i.e. dd) if they are morbidly curious,
>         provided that
>         the backing storage is fast enough to keep up with the vcpus
>         that xen is
>         presenting.
>         
>         BTW, removing xen-devel from the CC list, this is strictly a
>         question
>         for xen-users.
>         
>         Cheers,
>         --Tim
>         
>         
> 
> _______________________________________________
> 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