We use them all the time, and openstack in one version actually broke them on us. (I wrote and contributed a unit test so it shouldn't happen again.)
Use case: You have two external networks. 1. Internet - One that's directly connected to the internet. 2. One that is a private network space and is available to the whole cloud. Each tenant gets a router on each network (two routers total), defaulting to the internet one, and the subnet has a static routing rule to make the external private net route to the right neutron router. The user can then add floating ip's to one or both of the networks making the vm available on that network. If the service is internet facing, they can just put that type of floating ip on. If they want to share it with the other tenants but not with the internet, they just put that type of floating ip on. We don't have many ip's on the internet side, so having it split like this allows us to conserve ip's. Make sense? Thanks, Kevin ________________________________ From: Rubab Syed [rubab.sye...@gmail.com] Sent: Tuesday, March 08, 2016 2:20 PM To: openstack-operators@lists.openstack.org Subject: [Openstack-operators] [openstack-operators][neutron] use-case for multiple routers within a tenant Hi all, I am trying to get a general understanding of OpenStack networking. Can someone please point out a simple use-case for using multiple routers within same tenant? Thanks, Rubab
_______________________________________________ OpenStack-operators mailing list OpenStack-operators@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
