Description of problem:
After deploying a node as a controller node by associating it with the quickstack::controller puppet class, the horizon web server shows an error page immediately upon logging in.
Steps to Reproduce:
1. Deploy a new controller node through the foreman UI.
2. Log in to the Horizon dashboard on the controller node.
Error page is displayed.
An openstack dashboard is displayed.
I am able to get the Horizon web page to load only by deleting the keystone cinder endpoint, e.g.:
/usr/bin/keystone --token 3789f6674b5305851e1ba51cdc4cca3e --endpoint http://127.0.0.1:35357/v2.0/ endpoint-list
# delete the one with port 8776
usr/bin/keystone --token 3789f6674b5305851e1ba51cdc4cca3e --endpoint http://127.0.0.1:35357/v2.0/ endpoint-delete 7f8821476aff448f892f2b8b576daadc
More info here: http://pad.engineering.redhat.com/fm-horizon-cinder-connection-refused
Possibly same issue:
This issue has a workaround as Crag describes, so I don't see this as a blocker for snap2 or RC, but it needs to be addressed for GA.
Regardless Crag is investigating right now.
IIUC Crag has this fixed and in a build
QA ack for inclusion in the RC.
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.