James, 1. `openstack overcloud ceph deploy` is now supposed to configure NTP [0] as documented [1] 2. By default it uses pool.ntp.org which is blocked by Red Hat IT 3. Thus, you must override the NTP servers by using `openstack overcloud ceph deploy --ntp-server clock.redhat.com` 4. IR has been patched to support this [2] as have our phase 1/2 jobs. Any additional tests you have need to be updated accordingly. The experience you are having is expected and can be resolved as described above. Needinfo: please confirm the instructions above resolve your problem, and if they do, then please close "not a bug". John [0] https://review.opendev.org/q/topic:deployed_ceph_ntp [1] https://bugzilla.redhat.com/show_bug.cgi?id=2203793 [2] https://review.gerrithub.io/c/redhat-openstack/infrared/+/553732
Needinfo requested >1 week ago but still not answered. As per my last update it should be clear what to do. I'm going to close this bug. If you need to re-open it that's fine but please provide the necessary information if you do.
I've been working on how to configure this within out particular testcase (IPv6 with a l3appliance). It's a matter of finding where our testcase needs the correct overrides. Doing a manual edit and then re-running the Ceph deploy worked, and the Overcloud looks to be correct as well. So for now I think we're good.