Things to check to automatically determine if Postgresql Cluster already exists

[PatrickEPG]

Hi,
we want to fully automate the deployment of our Cluster in a Gitlab Pipeline and therefore want to populate the inventory dynamically.
Since we are currently planning on running deploy_pgcluster in every Pipeline, we need to set postgresql_exists accordingly when we populate the inventory.

I saw that Autobase verifies if postgres exists/does not exist in "automation/roles/pre-checks/tasks/patroni.yml", but this only applies to the initdb bootstrapping method.

Is this just irrelevant when using a different bootstrapping method?

Are there other things we need to check before setting the postgresql_exists variable automatically?

Thank you

[vitabaks]

postgresql_exists is a variable used to check for existing data. When a cluster is deployed, the data directory is recreated as empty to prevent data loss in scenarios where users are converting an existing PostgreSQL instance into a cluster.

This variable is only relevant during the initial deployment using the deploy_pgcluster.yml playbook.

[vitabaks]

By the way, in CI, for already deployed clusters, next we run the config_pgcluster.yml playbook

Doc: https://autobase.tech/docs/management/gitops

[PatrickEPG]

Okay so I might have misunderstood the logic then. So it is intended to run the deploy_pgcluster.yml once for initial deployment and from there on only run the config_pgcluster.yml ?

[vitabaks]

Yes, deploy_ playbook is being executed for initial deployment.

[PatrickEPG]

alright thank you.