openstack/openstack-ansible
Code Issues Proposed changes

[doc] Update documentation for galera cluster recovery

Browse Source 
Prior doc was misleading on a way how to re-bootstrap galera cluster.
It's now updated with relevant iinformation, that should trigger bootstrap
and rolling restart of the cluster.

Closes-Bug: #1927148
Change-Id: I0aee053cc6fcd6c8eb2261c20dba4e338ff61735
This commit is contained in:
Dmitriy Rabotyagov 2024-02-02 11:26:54 +01:00
parent 48eed425ef
commit b250e3b028
1 changed files with 11 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
doc/source/admin/maintenance-tasks/galera.rst
View File

@ -149,14 +149,23 @@ one of the nodes.
Galera cluster recovery Galera cluster recovery
----------------------- -----------------------
Run the ``galera-install`` playbook using the ``galera-bootstrap`` tag Run the ``galera-install`` playbook using the ``galera_force_bootstrap`` variable
to automatically recover a node or an entire environment. to automatically recover a node or an entire environment.
#. Run the following Ansible command to show the failed nodes: #. Run the following Ansible command to show the failed nodes:
.. code-block:: shell-session .. code-block:: shell-session
# openstack-ansible galera-install.yml --tags galera-bootstrap # openstack-ansible galera-install.yml -e galera_force_bootstrap=True --tags galera_server-config
You can additionally define a different bootstrap node through
``galera_server_bootstrap_node`` variable, in case current bootstrap node is in
desynced/broken state. You can check what node is currently selected for
bootstrap using this ad-hoc:
.. code-block:: shell-session
root@aio1:/opt/openstack-ansible# ansible -m debug -a var="groups['galera_all'][0]" localhost
The cluster comes back online after completion of this command. If this The cluster comes back online after completion of this command. If this
fails, please review `restarting the cluster`_ and `recovering the primary fails, please review `restarting the cluster`_ and `recovering the primary