Molecule - ec2 delete other instance present in a VPC

[shubham823]

Molecule instance before spinning up a instance it deletes all the other instances in present in a vpc which is a major bug.

[shubham823]

- name: Destroy ephemeral EC2 instances
  ec2_instance:
    profile: "{{ item.aws_profile | default(omit) }}"
    region: "{{ item.region | default(omit) }}"
    instance_ids: "{{ instance_config | map(attribute='instance_ids') | flatten }}"
    state: absent
  loop: "{{ platforms }}"
  loop_control:
    label: "{{ item.name }}"
  register: ec2_instances_async
  async: 7200
  poll: 0

it can be fixed by adding one line.

- name: Destroy ephemeral EC2 instances
  ec2_instance:
    profile: "{{ item.aws_profile | default(omit) }}"
    region: "{{ item.region | default(omit) }}"
    instance_ids: "{{ instance_config | map(attribute='instance_ids') | flatten }}"
    state: absent
  loop: "{{ platforms }}"
  loop_control:
    label: "{{ item.name }}"

code start

  when: instannce_config != '{}'

code end

  register: ec2_instances_async
  async: 7200
  poll: 0

[apatard]

Looks like a dup of #121 ?

[danielpodwysocki]

Can this be closed off since it looks like it's been fixed already?

#170

I wandered into here because I ran into it as well, but my destroy playbooks just needed updating

A fix for anyone still encountering it would be to update the plugin and delete -> re-generate the destroy scenarios under molecule/ in their role's directory.