Open Stack

Sam, Thanks, But I failed to install two compute nodes if tacker is enabled, I’m not sure if tacker depends on Netron L3, can you share local.conf for two compute nodes case with tacker enabled.

From: Sam Hague [mailto:shague at redhat.com]
Sent: Friday, November 20, 2015 9:19 PM
To: Yang, Yi Y <yi.y.yang at intel.com>
Cc: openstack-dev at lists.openstack.org; Flavio Fernandes <ffernand at redhat.com>; Tim Rozet <trozet at redhat.com>; Andre Fredette <afredette at redhat.com>
Subject: Re: [openstack-dev] [Neutron] How are Tenant VMs' traffic routed to Service VMs

Yi,

yes, you just create a router and connect the two networks to it. The router will ensure traffic works between the two networks/subnets. Add the router and then just add each subnet to the router. Something like the below:

neutron net-create vx-net --provider:network_type vxlan --provider:segmentation_id 1500
neutron net-create vx-net2 --provider:network_type vxlan --provider:segmentation_id 1501
neutron subnet-create vx-net 10.100.5.0/24<http://10.100.5.0/24> --name vx-subnet --dns-nameserver 8.8.8.8
neutron subnet-create vx-net2 10.100.6.0/24<http://10.100.6.0/24> --name vx-subnet2 --dns-nameserver 8.8.8.8

neutron net-create ext-net
neutron router-interface-add ext-rtr vx-subnet
neutron router-interface-add ext-rtr vx-subnet2

Thanks, Sam

On Fri, Nov 20, 2015 at 2:52 AM, Yang, Yi Y <yi.y.yang at intel.com<mailto:yi.y.yang at intel.com>> wrote:
Hi, folks

I'm trying tacker to start some service VMs as Service Function VNF by "heat" tenant user, service VMs have special Neutron net & subnet, other common tenant VMs will have their own Neutron net & subnet, my question is how to route the traffic to service VMs in Openstack environment, DVR or router?

I integrated Opendaylight and used Opendaylight ML2 driver (https://github.com/openstack/networking-odl), in that case, I used its L3 routing plugin instead of Neutron L3, I also integrated ovsdb, from ovsdb perspective, ARP response and L3 routing are done by openflow tables, so can openflow tables do the same thing to routing the traffic between tenant VMs and service VMs?

Network topology looks like the below diagram.


+--------------------------------------------------------------------------+                            +------------------------------------------------------------------------------+
|Compute Node 1                             |                            |Compute Node 2                               |
|                                           |                            |                                            |
|                                           |                            |                                            |
|                                           |                            |                                            |
|                                           |                            |                                            |
|                                           |                            |                                            |
|       +--------------------+      +---------------------+   |                            |       +--------------------+       +-----------------+    |
|       |Tenant VM1  |      |Service VMx  |   |                            |       |Tenant VM2  |       |Service VMy|    |
|       | 10.0.0.3     |      | 11.0.0.3    |   |                            |       | 10.0.0.4     |       | 11.0.0.4   |    |
|       |            |      |            |   |                            |       |            |       |          |    |
|       |            |      |            |   |                            |       |            |       |          |    |
|       |            |      |            |   |                            |       |            |       |          |    |
|       +-------eth0-------+      +--------eth0------+   |                            |       +-------eth0-------+       +------eth0-----+    |
|              |                   |         |                            |              |                   |         |
|              |                   |         |                            |              |                   |         |
|             tap0                tap1       |                            |             tap0                tap1        |
|              |        ovs        |         |                            |               |        ovs        |         |
|              +-------------br-int------------+         |                            |               +------------br-int------------+         |
|                        |                   |                            |                         |                  |
|              +----------------+---------------+         |                            |               +----------------+---------------+        |
|              |                   |         |                            |               |                   |        |
|            VxLAN1          VxLAN-gpe1      |                            |            VxLAN1            VxLAN-gpe1   |
|                                            |                            |                                           |
|-------------------------------------eth1---------------------------------+                            + ---------------------------------------eth1-----------------------------+
            192.188.50.3|                                                                           | 192.168.50.4
                       +===========================================================================+

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20151123/d8fc529f/attachment.html>