If it's non-devstack setup, you need to set "local_ip" other_config in ovsdb data base for tunnel creation. Also for l3 services (router) you need to set ovsdb.l3.fwd.enabled=yes in custom.properties

Greetings! For the folks following the series on ODL OVSDB net-virt, I finally wrapped up the page on how we are handling one-to-one nat in Lithium:

Hi, As I understand, the commit for BR 3378 would 1) Take away the need to add the gateway mac in the /etc/custom.properties. 2) Help verify the VM’s ability to reach the external

Thanks for the information, Sam. I look forward to discussing this in tomorrow's meeting. Regards, Ryan Goulding

Ryan, Daya, these items would be part of the work Anil and Flavio are driving for the clustering,persistence and ha in Be. There are high-level cards in the Be Trello board. They are hoping to

Yi, it is only partially working in your setup. Your ovsdb nodes are connected to ODL and the pipeline flows are created which is good. You can see that in the final dump-flows in your output. But

Hi Ed, Should we make a trello card for this? If no one has started work on this, I would be interested in picking this up. Thanks, Ryan Goulding

i had a related question, for switch restarts. do we rely on ovs persistence to save the config, or do we always push down the controller's version of the config to the switch? if former, then do we

Hi Yi The flows in br-int are configured by ODL. OpenStack normally uses tags instead of flows to route the packets. However please check the service plugin set for L3 services in Neutron config file

Hi, All I followed https://wiki.opendaylight.org/view/OpenStack_and_OpenDaylight to integrate Openstack and SFC. I saw br-int was created by ODL. [root@localhost ~(keystone_admin)]# ovs-vsctl

Hi Eric, Did you try to to use "dhclient ethX" on the VM or if you are using cirros image use "udhclient ethX"? It will send the dhcp request again to get the ip address. If this works, then it's

I’ve been noticing, when using Netvirt, that the Openflow connectivity to the ‘private’ network created by devstack is typically (always?) not set up completely. As a result, VMs instantiated

Added neutron-dev in the loop.

Hi OVSDB, there is a job [0] that is supposed to be running some robot tests [2]. However, we noticed that the job is failing and does not get to the robot part at all. I know there was a lot of

+1. Regards, Ryan Goulding

So we probably also need to get one for retrying connections.. because currently if the ovsdb node is not available or reachable when we configure a connection, we never retry, and if it goes away

https://trello.com/c/bOrmGbXQ/46-resync-persisted-config-to-ovsdb-correctly-on-restart-of-controller is a trello card that Eric has taken involving controller restart for OVSDB southbound. This is

Guys, Is anyone working on these issues with OVSDB SB? Ed

I can get rid of those branches in gerrit, let me do that now.

yes, make sense. How and where is this hash kept? I think whatever hash it has now is fine, except for the gerrit I propose to be reverted [1]. @Kyle: agree? Yes. Can you give me more concrete steps