stopping ec2 spot instances

Archives Ansible Project
1

I’m writing an ansible demo, showing how to provision Riak clusters, and I want to make it super self contained.

With that in mind, I’ve been playing with the ec2.py inventory script, and the ansible ec2 module.

Starting the nodes works great (if anyone is interested, I’m happy to share the code, or perhaps just make the repo public)

But, I’m having a problem with stopping the nodes.

I have those vars available:

TASK: [List tagged instances] *************************************************
ok: [127.0.0.1] => {
“groups["tag_instance_type_ansible_demo"]”: [
“54.171.160.194”
]
}

TASK: [everything for that one host] ******************************************
ok: [127.0.0.1] => {
“hostvars["54.171.160.194"]["ec2_id"]”: “i-f28793b0”
}

The problem is that in order to stop and ec2 instances, I need to get at the ec2_id, in order to feed it to the ec2 module (which stops nodes based upon the instance id, not the ip address.

tasks:

  • name: shut down all the spot instances
    local_action:
    module: ec2
    region: eu-west-1
    wait: no
    instance_ids: ‘{{ instance_ids }}’
    state: absent

I can’t figure out how to munge this into a usable list, has anyone else experienced this before?

Bryan

2

Can you get this using the “ec2_facts” module or are you using the dynamic inventory script?

3

This is how I got something similar working in order to terminate instances. You will need to change the state and I’m using the inventory script as my hosts file. I’m sure it can be cleaned up. In this case, I’m targeting ec2 instances with the Nginx_Server tag.

`

Connect to the node and gather facts,

including the instance-id. These facts

are added to inventory hostvars for the

duration of the playbook’s execution

Typical “provisioning” tasks would go in

this playbook.

  • hosts: tag_Name_Nginx_Server
    gather_facts: True
    user: root
    sudo: True
    tasks:

fetch instance data from the metadata servers in ec2

  • ec2_facts:

show all known facts for this host

#- debug: var=hostvars[inventory_hostname]

just show the instance-id

  • debug: msg=“{{ hostvars[inventory_hostname][‘ansible_ec2_instance_id’] }}”

Using the instanceid, call the ec2 module

locally to remove the instance by declaring

its state is “absent”

  • hosts: tag_Name_Nginx_Server
    gather_facts: True
    connection: local
    tasks:
  • name: destroy all instances
    ec2: state=‘absent’
    instance_ids={{ item }}
    wait=true
    with_items: hostvars[inventory_hostname][‘ansible_ec2_instance_id’]

`