Software Factory 3.8 Zuul/Nodepool Update

# Backup your Software Factory data
ansible-playbook /var/lib/software-factory/ansible/sf_backup.yml

# Install this specific sf-config package
yum install sf-config-3.8.9-4

# Ensure to deactivate the automatic restart of Zuul and Nodepool
echo "disable_zuul_autorestart: true" >> /etc/software-factory/custom-vars.yaml
echo "disable_nodepool_autorestart: true" >> /etc/software-factory/custom-vars.yaml

# Stop all Zuul/Nodepool services (adapt the command if services are running on multiple nodes)
systemctl stop zuul-scheduler zuul-merger zuul-web zuul-executor zuul-fingergw \
 nodepool-launcher nodepool-builder

# A Zuul SQL database migration need to be performed by Zuul at startup and we have noticed
# an issue with our DBs content that prevent the migration to success. To fix it, run:
podman exec -it mysql bash -c 'mysql -uroot -p$MYSQL_ROOT_PASSWORD zuul -e "delete from zuul_buildset where (CHAR_LENGTH(oldrev) > 40 OR CHAR_LENGTH(newrev) > 40 OR CHAR_LENGTH(patchset) > 40);"'

# Delete the Zuul Ephemeral state from Zookeeper
zuul_wrapper delete-state

# Then run the sf-config upgrade command
sf-config --upgrade

# Re-create the zuul-scheduler container based on the updated container image
podman rm zuul-scheduler; /usr/local/bin/container-zuul-scheduler.sh; rm /var/lib/software-factory/versions/zuul-scheduler-updated

# Now start the zuul-scheduler and ensure that the database migration has been performed without issue.
systemctl start zuul-scheduler
tail -f /var/log/zuul/scheduler.log

# Perform a full restart of Zuul
ansible-playbook /var/lib/software-factory/ansible/zuul_restart.yml

# Then restart all Nodepool services
ansible-playbook /var/lib/software-factory/ansible/nodepool_restart.yml