Yes, to clarify, on train we are using master and on rocky we are using stable/rocky. Train and Stein exhibit the same behavior, Rocky works successfully.
With regard to the logs and current bug report:
Yes, we are using kolla-ansible stable/stein deploying the official stein images.
Yes, currently we are deploying haproxy:children to (network) not (control).
Once we get past the keystone issue we intend to change haproxy:children over to (control) but that seems to cause a different failure. However, I assumed it would be easier to figure out the keystone issue if I simplified the configuration.
Hi,
Yes, to clarify, on train we are using master and on rocky we are using stable/rocky. Train and Stein exhibit the same behavior, Rocky works successfully.
With regard to the logs and current bug report:
Yes, we are using kolla-ansible stable/stein deploying the official stein images.
Yes, currently we are deploying haproxy:children to (network) not (control).
Once we get past the keystone issue we intend to change haproxy:children over to (control) but that seems to cause a different failure. However, I assumed it would be easier to figure out the keystone issue if I simplified the configuration.
Sorry for any confusion.