Thank you all very much for the replies!

Gene

On 14-05-13 06:37 PM, Ruben S. Montero wrote:
> Yes, there maybe a problem. There won't be nothing to stop the
> scheduler from allocating the VM if it's resumed. That's the reason
> for accounting the capacity even if you are not actually using it,
> you'll be granted with it as soon you request it (and its available). 
>
> So you may be not making a "complete" use of the resource them but you
> have a "reservation" for them, especially for SUSPENDED/POWER OFF. But
> also for HOLD, PENDING, STOP, that will potentially interfere with
> other users when the VM is resumed.
>
> Now we could check the quota when the VM is moved to PENDING. For
> example, OpenNebula won't let you resume the VM. 
>
> However, the other quotas should be also modified to follow the same
> convention. For example, number of IP's leased from a network. The
> main problem I see here is that would be conflicting with other
> metrics like (number of VMs). If we only limit ACTIVE VMs a DoS attack
> could be done easily by creating VMs in STOP (e.g. size of checkpoint
> files is implicitly controlled by the MEMORY quota). If we preserve
> the quota as in 4.6, then the MEMORY quota for ACTIVE VMs wouldn't be
> useful. 
>
> Maybe is not "fair" in all cases (I believe that only stop VMs are a
> problem, and that operation is not exposed for the end-users, i.e.
> cloud view ) but it is consistent for all the quotas...
>
> Cheers
>
> Ruben
>
>
>
>
> On Tue, May 13, 2014 at 11:15 PM, Jon <[email protected]
> <mailto:[email protected]>> wrote:
>
>     How would that affect scheduling? For instance I have a vm that
>     only comes on at certain times of the day.
>
>     I think this is a good change too, I'm just curious if there are
>     any repercussions to the change.
>
>     Thanks,
>     Jon A
>
>     On May 13, 2014 2:04 PM, "kiran ranjane" <[email protected]
>     <mailto:[email protected]>> wrote:
>
>         +1 vote for this
>
>         Thanks
>
>         Kiran Ranjane
>
>
>         On Tue, May 13, 2014 at 3:05 PM, Stefan Kooman <[email protected]
>         <mailto:[email protected]>> wrote:
>
>             Quoting Carlos Martín Sánchez ([email protected]
>             <mailto:[email protected]>):
>             > Hi,
>             >
>             > On Fri, May 2, 2014 at 10:51 PM, Liu, Guang Jun (Gene) <
>             > [email protected]
>             <mailto:[email protected]>> wrote:
>             >
>             > >  Hi there,
>             > >
>             > > Is there a way to config or twist OpenNebula so that
>             quota management
>             > > (e.g. RAM) counts resources of active VMs only?
>             > >
>             > > I found the quota management (RAM per user) counts all
>             the VMs which
>             > > created by a user. We found if the quota only counts
>             the resources from the
>             > > active VMs, that means more sense. E.g., I have 16RAM
>             quota. I create one
>             > > VM w/ 16G RAM. Then I save/stop my VM. The system
>             should allow me to create
>             > > another VM w/ 16G RAM if quota management counts only
>             active VMs.
>             > >
>             > > Thanks!
>             > > Gene
>             > >
>             >
>             > There isn't a way to configure that. It would require to
>             modify the core.
>             >
>             > But I think it makes sense. So I return the question to
>             the community:
>             > ¿Should we consider this change in the quota behaviour?
>
>             I vote for this change. Your not actually consuming
>             resources on the
>             hypervisor(s) for not having running vm's ...
>
>             Gr. Stefan
>
>             --
>             | BIT BV  http://www.bit.nl/        Kamer van Koophandel
>             09090351
>             | GPG: 0xD14839C6                   +31 318 648 688
>             <tel:%2B31%20318%20648%20688> / [email protected]
>             <mailto:[email protected]>
>             _______________________________________________
>             Users mailing list
>             [email protected] <mailto:[email protected]>
>             http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>
>
>         _______________________________________________
>         Users mailing list
>         [email protected] <mailto:[email protected]>
>         http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>
>     _______________________________________________
>     Users mailing list
>     [email protected] <mailto:[email protected]>
>     http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>     -- 
>     -- 
>     Ruben S. Montero, PhD
>     Project co-Lead and Chief Architect
>     <http://lists.opennebula.org/listinfo.cgi/users-opennebula.org>
>     OpenNebula - Flexible Enterprise Cloud Made Simple
>     
> <http://lists.opennebula.org/listinfo.cgi/users-opennebula.org>www.OpenNebula.org
>     <http://www.OpenNebula.org> | [email protected]
>     <mailto:[email protected]> | @OpenNebula
>
>
>
> _______________________________________________
> Users mailing list
> [email protected]
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

_______________________________________________
Users mailing list
[email protected]
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Reply via email to