starlingx/docs
Code Issues Proposed changes
docs/doc/source/backup/kubernetes/node-replacement-for-aiominussx-using-optimized-backup-and-restore-6603c650c80d.rst
Elisamara Aoki Goncalves fd149454f2 Add requirement for node replacement 
Closes-bug: 2002434

Signed-off-by: Elisamara Aoki Goncalves <elisamaraaoki.goncalves@windriver.com>
Change-Id: I0ca7643e754984489b4bfcaeba37b5d0478f455c
2023-01-10 15:33:17 -03:00

1.4 KiB
Raw Blame History

AIO-SX - Restore on new hardware

For Node Replacement with configured ceph storage backend it is not possible to restore ceph data, since OSD disks will not be present to recover. The flag wipe_ceph_osds=true should be set in this case when running the optimized restore playbook.

Note

The Node replacement can only be performed with equivalent hardware that has the same physical attributes and configuration.

The improved backup and restore supports Node Replacement, by the following method.

Restore Playbook Parameter to Specify Node Replacement

An optional parameter is provided to the restore playbook with the new management MAC address:

-e replacement_mgmt_mac <mgmt_mac>

This parameter is accepted only for controller-0.

For example:

~(keystone_admin)]$ ansible-playbook /usr/share/ansible/stx-ansible/playbooks/restore_platform.yml -e "initial_backup_dir=/home/sysadmin" -e "ansible_become_pass=St8rlingX*" -e "admin_password=St8rlingX*" -e "backup_filename=localhost_platform_backup.tgz" -e "restore_mode=optimized" -e "restore_registry_filesystem=true" -e "replacement_mgmt_mac=a1:a2:a3:a4:a5:a6"