[web/task split] fix scale down bug

[thedoubl3j]

SUMMARY

fixes #1255

While working on cleanup, the scale down playbook didn't have the new naming conventions for the pods and would currently scale down nothing during any type of db upgrade etc.

This adds a simple list loop to scale down both pods during any db upgrade/migration/configuration in the operator.

ISSUE TYPE

• Bug, Docs Fix or other nominal change

ADDITIONAL INFORMATION

test case
To test this properly, we will need to either provoke the rescue block or do a db migration or upgrade. I think there is also a way to provoke it if there is no DB specified but I need to investigate this.

If anyone has any pointers or thoughts on how we can approach this, please share as I am running low on ideas.

[rooftopcellist]

@thedoubl3j It looks like there is an entry in roles/restore that also needs to be addressed:

• awx-operator/roles/restore/tasks/postgres.yml

  Lines 49 to 65 in 9c788ba

  	- name: Check for presence of AWX Deployment
  	k8s_info:
  	api_version: v1
  	kind: Deployment
  	name: "{{ deployment_name }}"
  	namespace: "{{ ansible_operator_meta.namespace }}"
  	register: this_deployment
  	- name: Scale down Deployment for migration
  	k8s_scale:
  	api_version: v1
  	kind: Deployment
  	name: "{{ deployment_name }}"
  	namespace: "{{ ansible_operator_meta.namespace }}"
  	replicas: 0
  	wait: yes
  	when: this_deployment['resources'] | length

[thedoubl3j]

@rooftopcellist take a look when you get a sec, I elected to check for a task deployment rather than just awx (doesn't exist anymore) or both since if one is there, the other will be as well.