Collection management and automation

[Michael Gutteridge]

We are planning a GCS deploy where we (the IT support folks) create the mapped collection that our labs would then use to create guest collections for their specific needs.  The idea is that we put some useful guardrails around the data they own, but they still retain control over the actual sharing of data.

We'd like to put in automation to create and manage that base mapped collection- partially 'cos we're lazy, mostly so that those collections are created correctly.  I'm curious if anyone else is automating collection management, if so, how?  A configuration management tool (Ansible, Chef, Puppet) or some kind of home-grown script?  Anyone using something like Terraform or OpenTofu?

We are using Ansible to manage the configuration of hosts and initially looked at using Ansible for managing collections.  I saw this thread and started working with the Ansible module Yuriy had developed.  It was looking promising, but the problem I've run into is that since the collection ID isn't available until after the collection is created, there needs to be some mechanism to store that collection ID with the Ansible variables/tasks necessary to manage it.

The Terraform/OpenTofu route is appealing: persistent storage of the infrastructure state is built-in and this task- provisioning infrastructure- is something it's very good at.  The downside for us is that we are minimal Terraform users and golang isn't a strength.  It'd be kind of a one-off in our environment.

Anyway, I'm curious as to what others are doing here.  Or maybe I'm thinking about this all wackadoodle and we should revisit how we manage this service.

Thanks!

 - Michael