Issue

• Director nova-api does not respond to metadata requests causing a scaleout to timeout and fail.
• Openstack scale out fails due to metadata server unreachability in RHOSP even when nova-api are up and running.
• below logs are found in controller node :

Sep 07 16:45:44 hostname.example.com os-collect-config[5094]: HTTPConnectionPool(host='169.254.169.254', port=80): Read timed out. (read...=10.0)

• We see following in newly added compute node after logging in (scale out process timed out)

Jul 13 03:08:55 hostname.example.com os-collect-config[3215]: /var/lib/os-collect-config/local-data not found. Skipping
Jul 13 03:08:55 hostname.example.com os-collect-config[3215]: No local metadata found (['/var/lib/os-collect-config/local-data'])
Jul 13 03:09:25 hostname.example.com os-collect-config[3215]: /var/lib/os-collect-config/local-data not found. Skipping
Jul 13 03:09:25 hostname.example.com os-collect-config[3215]: No local metadata found (['/var/lib/os-collect-config/local-data'])
Jul 13 03:09:55 hostname.example.com os-collect-config[3215]: /var/lib/os-collect-config/local-data not found. Skipping
Jul 13 03:09:55 hostname.example.com os-collect-config[3215]: No local metadata found (['/var/lib/os-collect-config/local-data'])
Jul 13 03:10:28 hostname.example.com os-collect-config[3215]: 500 Server Error: Internal Server Error for url: http://169.254.169.254/latest/meta-data/