c2d1102574
We have been using ctlplane network by default for Keystone admin endpoint. The intention was to let undercloud talk with the endpoint to create resources like endpoints, but that requirement is not true, since we use tripleo-ansible, which relies on public endpoint, to create keystone resources. This change updates the default to internal api network. The ctlplane network, which is the provisioning network, usually has no redundancy, and we should not make any of overcloud functionalities rely on this network if there is no actual requirement to do so. Closes-Bug: #1959459 Change-Id: I931e4c3c8a081adb144d915410e0d38760466561 |
||
|---|---|---|
| .. | ||
| README.md | ||
| update-from-192_0_2-subnet.yaml | ||
| update-from-deployed-server-newton.yaml | ||
| update-from-ml2-ovs-from-rocky.yaml | ||
| update-from-overcloud-compute-hostnames.yaml | ||
| update-from-publicvip-on-ctlplane.yaml | ||
README.md
This directory contains Heat environment file snippets which can be used to ensure smooth updates of the Overcloud.
Contents
update-from-keystone-admin-internal-api.yaml To be used if the Keystone Admin API was originally deployed on the Internal API network.
update-from-publicvip-on-ctlplane.yaml To be used if the PublicVirtualIP resource was deployed as an additional VIP on the 'ctlplane'.
update-from-deloyed-server-newton.yaml To be used when updating from the deployed-server template from Newton.