Thank you for socializing this here. However, the best course to a resolution in an OpenStack Charm is generally through raising a bug. That will allow us to triage it and track its status and log attachments as it progresses, in alignment with the other OpenStack Charms.
Here is an extract from a known-good bundle which illustrates shared-db space usage on the openstack-dashboard charm. Please look this over and raise a bug if the issue persists. Thanks again!
@beisner Thanks for that … is an interface, not a network endpoint binding…
If I’m picking up what you are laying down, you are telling me that we can bind regular relation endpoint interfaces to spaces as well as network-binding endpoints?
The “shared-db” endpoint that you have referenced is not actually a network-binding, it’s a “requires” interface endpoint that you are binding to the space (I think).
If this is the case, then what is the point of network-bindings?
Hey dear @beisner , your replies clear many points to me for openstack bundle deployment. but can you please explain this to me:
in the openstack bundle example there is one or two vip assigned to each service with hacluster as below:
first can you specify each subnet related to which network space? I mean 10.244.40.81 is from oam, admin, or public space. and same for 192.168.33.3.
for my case I deployed the same bundle but the dashboard default binding is to oam space which is the only space that have access to the internet for deploying (while in the example the default binding is to the public space) and I am not able to access the dashboard from any vip.
then I tried to bind public endpoint of dashboard to public-api space and here I were able to access but can’t do anything from the admin tab in the left menu of openstack dashboard.