devops: disks are being accumulated for temporal and long-live instances

[gustavovalverde]

Describe the issue or request

We need to cleanup disks that are being left behind when replacing an old instance with a new one

Expected Behavior

On deletion, existing disks should be deleted if we're mounting an cached state, and be kept if we're re-mounting the same disk

Current Behavior

We don't have a workflow or logic that handles this differentiation

Possible Solution

A cleanup job

Additional Information/Context

No response

Is this happening on PRs?

No

Is this happening on the main branch?

Yes