Refactor config_pgcluster.yml, add the leader identification and backup settings

[vitabaks]

• Add the leader (Primary) identification and dynamic inventory for the config_pgcluster.yml playbook.

  • Previously, our approach relied on the "master" group in the inventory to designate tasks such as user or database creation. This presented challenges when the designated host in the inventory transitioned from a leader to a replica (e.g., post-switchover). Such transitions resulted in errors since database object creation mandates execution on the primary (leader) node.

• Add the pgBackRest and WAL-G roles into the config_pgcluster.yml playbook.

  • Previously, backup configurations were absent in the 'config_pgcluster.yml' playbook. This posed an inconvenience: any modification to backup configurations or schedules necessitated the execution of 'deploy_pgcluster.yml'. This wasn't optimal since 'deploy_pgcluster.yml' is primarily intended for the initial cluster deployment.