[openstack-dev] [neutron] openvswitch integration

Andreas Scheuring scheuran at linux.vnet.ibm.com
Thu Oct 2 07:20:12 UTC 2014


Hi together, 
I'm wondering why ovs was integrated into openstack in the way it is
today
(http://docs.openstack.org/grizzly/openstack-network/admin/content/under_the_hood_openvswitch.html)

Especially I would like to understand
- why does every physical interface have its own bridge? (I guess you
could also plug it directly into the br-int)
- and why does the br-int use vlan separation and not directly the
configured tenant-network-type separation (e.g. vxlan or something
else)? Tagging a packet with the internal vlan and then converting it to
the external vlan again looks strange to me in the first place.

It's just a feeling but this surely has impact on the performance. I
guess latency and cpu consumption will surely go up with this design.
Are there any technical or historical reasons for it? Or is it just to
reduce complexity?

Thanks!


-- 
Andreas 
(irc: scheuran)






More information about the OpenStack-dev mailing list