chore(tenants): tenant cleanup

[lusergit]

Closes #504.
Manages resource cleanup on tenant deletion, deleting base images, ephemeral images and system models images on remote storage.

[coveralls]

Pull Request Test Coverage Report for Build 83a215a0aeb5ccb5c8de527747f9eeb256e6f32a-PR-775

Details

• 0 of 0 changed or added relevant lines in 0 files are covered.
• 2 unchanged lines in 1 file lost coverage.
• Overall coverage increased (+0.2%) to 78.387%

---

Files with Coverage Reduction	New Missed Lines	%
lib/edgehog/tenants/tenant/tenant.ex	2	60.0%

Totals
Change from base Build e25b04d0901a73c89d3b135e2715dbe5e3bc9fc4:	0.2%
Covered Lines:	1244
Relevant Lines:	1587

---

💛 - Coveralls

[davidebriani]

Keep in mind that inside the Ash.Changeset.after_action logic the multitenant resources are already deleted, thanks to the cascade deletes and the tenant being deleted.
We could either:

• verify that Ash automagically calls the destroy actions on resources instead of relying on the cascade deletes, and we add Changes to the destroy actions that try to cleanup the resources from external systems.
• read the resources before the after_action callback, and then call directly the modules (e.g. @ephemeral_image_module.delete) to cleanup the external systems.

[lusergit]

For the first point, because of the way we set up the deletion (using ash_postgres references) ash does not automatically call the destroy actions of associated resources (see ash_postgres documentation). I think the second route is our best option, as it allows to handle the deletion directly in the resources (with a before_action) that triggers the external sotrages behavior only after the transaction if successful.

[lusergit]

I updated the commit with this approach, the resources get deleted even on external storage (see tenants_test)