openstack team mailing list archive

[Gary Kotton <gkotton@xxxxxxxxxx>]

On 01/22/2013 07:02 PM, Jian Wen wrote:
> On 2013?01?22? 10:56, Balamurugan V G wrote:
> > Hi,
> >
> > In an Openstack deployment using Quantum, the compute nodes access 
> > the public/external network via the network node (as per the diagram 
> > http://docs.openstack.org/trunk/openstack-network/admin/content/connectivity.html).
> >
> > So if I have a hundred compute nodes each running hundred instances, 
> > wouldn't the link to the network node choke? Also wouldnt the network 
> > node's resources be a bottleneck?
> >
> > Regards,
> > Balu
> >
> >
> > _______________________________________________
> > Mailing list:https://launchpad.net/~openstack
> > Post to     :openstack@xxxxxxxxxxxxxxxxxxx
> > Unsubscribe :https://launchpad.net/~openstack
> > More help   :https://help.launchpad.net/ListHelp
>  Quantum Multi-host DHCP and L3
> https://blueprints.launchpad.net/quantum/+spec/quantum-multihost

Yes. At the moment the network node is the bottleneck. There is a patch 
upstream in review that will solve this issue:
https://review.openstack.org/#/c/18216/

Hopefully this will be ready in the Grizzly cycle.

In addition to this it also provides HA for the L3 and DHCP agents.
Thanks
Gary

> --
> Jian Wen
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~openstack
> Post to     : openstack@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~openstack
> More help   : https://help.launchpad.net/ListHelp