Merge "Insufficient /tmp space to hold restore data during subcloud restore (r8, dsr8, r7, dsr7, r6, dsr6, r5, dsr5)"
This commit is contained in:
commit
79b5904bf8
@ -113,6 +113,23 @@ In this method you can run Ansible Restore playbook and point to controller-0.
|
||||
|
||||
~(keystone_admin)]$ ansible-playbook /localdisk/designer/jenkins/tis-stx-dev/cgcs-root/stx/ansible-playbooks/playbookconfig/src/playbooks/restore_platform.yml --limit |prefix|\_Cluster -i $HOME/br_test/hosts -e "ansible_become_pass=St0rlingX* admin_password=St0rlingX* ansible_ssh_pass=St0rlingX* initial_backup_dir=$HOME/br_test backup_filename= |prefix|\_Cluster_system_backup_2019_08_08_15_25_36.tgz ansible_remote_tmp=/home/sysadmin/ansible-restore"
|
||||
|
||||
.. warning::
|
||||
|
||||
If ``ansible_remote_tmp`` is not set, ``/tmp`` will be used. ``/tmp``
|
||||
can only hold 1GB.
|
||||
|
||||
An example of what happens when ``ansible_remote_tmp`` is not used:
|
||||
|
||||
.. code-block:: none
|
||||
|
||||
TASK [backup-restore/transfer-file : Transfer backup tarball to /scratch on controller-0] ***
|
||||
Wednesday 21 June 2023 13:59:28 +0000 (0:00:00.230) 0:00:51.283 ********
|
||||
fatal: [subcloud1]: FAILED! =>
|
||||
msg: |-
|
||||
failed to transfer file to /opt/platform-backup/subcloud1_platform_backup_2023_06_09_23_14_14.tgz /tmp/.ansible-sysadmin/tmp/ansible-tmp-1687355968.13-696694507261/source:
|
||||
|
||||
scp: /tmp/.ansible-sysadmin/tmp/ansible-tmp-1687355968.13-696694507261/source: No space left on device
|
||||
|
||||
- The ``ssl_ca_certificate_file`` defines a single certificate that
|
||||
contains all the ssl_ca certificates that will be installed during the
|
||||
restore. It will replace the
|
||||
|
Loading…
Reference in New Issue
Block a user