Improved way of detecting Galera cluster is up&running

There is a script in every node that checks correctly if the
Galera node has been correctly re-introduced in the cluster
as the previous check on the local port open had some
disadvantages, and does not wait for IST to complete.

This is an unclean cherry-pick because backup_and_restore in train
is called backup-and-restore. No extra changes implemented.

Change-Id: Id49274f5bade4af5b88b6354cfa092e4b041eb6e
(cherry picked from commit dc4e574ba1)
(cherry picked from commit 79f92e70ad)
This commit is contained in:
jlarriba 2022-10-17 14:20:11 +02:00
parent 0b7c84cd82
commit e628ec88a1
2 changed files with 2 additions and 15 deletions

View File

@ -65,13 +65,7 @@
- bar_create_recover_image
- name: Wait until pacemaker has Galera up&running
shell: |
set -o pipefail
ss -tunlp | grep ":3306 " | sed -e 's/.*\///'
register: mysql_result
retries: 300
until: mysql_result is search('mysqld')
delay: 5
shell: /var/lib/container-config-scripts/pacemaker_wait_bundle.sh galera galera-bundle Master
when:
- enabled_galera
tags:

View File

@ -21,13 +21,6 @@
- bar_create_recover_image
- name: Wait until pacemaker has Galera up&running
shell: |
set -o pipefail
ss -tunlp | grep ":3306 " | sed -e 's/.*\///'
register: mysql_result
retries: 300
until: mysql_result is search('mysqld')
delay: 5
when: tripleo_backup_and_restore_enabled_services.stdout is search('mysql^')
shell: /var/lib/container-config-scripts/pacemaker_wait_bundle.sh galera galera-bundle Master
tags:
- bar_create_recover_image