This is caused by the pool_must_be_enabled validation on deployments. Since it makes
sense to wrap up existing deployments after a pool has been disabled, I changed the
validation to only be on create. This also matches the current state of the similar
instance model validation.
Mainn