Openstack 13 enable Octavia.

Latest response

Hi Folks,
on OSP 13 LB should be already available. It is visible in horizon but it cannot be configured.
I have this instruction which does not work properly.

Any suggestion enabling octavia?

Thanks and Best Regards



Hi Niko,

So the /usr/share/openstack-tripleo-heat-templates/environments/services-docker/octavia.yaml should be the main environment file to include to enable octavia. When you say the instructions don't work, is there a particular point in the docs where it fails?

Hi Daniel, thanks for quick response. i did not use this file to enable octavia. In tried this: tripleo-common/workbooks/octavia_post.yaml tripleo-heat-templates/docker/services/octavia/octavia-deployment-config.yam

In the standard redhat documenation i did not find the information about the octavia.yaml you sentl. I will test it and let you know thanks BR Niko


I have same problem, when test create a loadbalance with octavia get this error :

| provisioning_status | ERROR |

Hi Pedro,

Any errors that pop up in /var/log/containers/octavia when you try to create a loadbalance?

I can't see any errors and nothing process running:

c46ee0659338 "kolla_start" 18 hours ago Up 18 hours (healthy) octavia_housekeeping


50bb7fd71054 "kolla_start" 18 hours ago Up 18 hours (healthy) octavia_health_manager


e4bcd79b7c8a "kolla_start"
18 hours ago Up 18 hours (healthy) octavia_api


3929b40a959c "kolla_start" 18 hours ago Up 18 hours (healthy) octavia_worker

Good morning - I am setting up a new RHOSP13 cloud for testing and need to deploy Octavia. My question - Did you add Octavia POST cloud deployment or just part of the normal cloud deployment? I ask because when I added it, I got errors that the LB networks did not exist and it failed out. It is possible that I made a mistake in a config file but trying to get an idea of the actual deployment flow.


It would be insightful to know the exact steps used to configure the LB when the provisioning_status went to ERROR. I have seen this behaviour in configuring a healthmonitor of TCP to an application port where it is not returning OK but rather just random text. However if I use a HTTP monitor for port 80 all works.

Hi, So here is the scenario in my all virtual setup comprises of; 1x Ctrl, 1x Cmp & 1x Ceph node

The strange thing happening here; deployment went all successful until the both Opendaylight and Octavia templates included at the same time during deployment. So my query here; is there any relation to both for enabling these services at the same time in Redhat Openstack Queens deployment or there's a specific order of calling them at the time of deployment or else.. any idea? Thanks

any clue or guidance will be appreciated in this regards pl!

@Nikolaos Raptis, you just need to add or call the template as mentioned before by @Daniel Macpherson

Guess I'm hitting this bug; so decided to roll the minor update and see the results.. lets see!