Ansible 1.3.2 doesn't seem to work with ec2 samples?

Archives Ansible Project
1

Hi,

I’ve just clean installed ansible 1.3.2 from rquillo’s PPA on my Ubuntu 12.04 system.

I created a playbook file (ansible-ec2.yml) with the following contents from the ec2 module documentation:

Launch instances, runs some tasks

and then terminate them

  • name: Create a sandbox instance
    hosts: localhost
    gather_facts: False
    vars:
    keypair: langfords
    instance_type: m1.small
    security_group: sg-1edd3571
    image: ami-c7c0d6b3
    region: eu-west-1
    tasks:
  • name: Launch instance
    local_action: ec2 keypair={{ keypair }} group={{ security_group }} instance_type={{ instance_type }} image={{ image }} wait=true region={{ region }}
    register: ec2
  • name: Add new instance to host group
    local_action: add_host hostname={{ item.public_ip }} groupname=launched
    with_items: ec2.instances
  • name: Wait for SSH to come up
    local_action: wait_for host={{ item.public_dns_name }} port=22 delay=60 timeout=320 state=started
    with_items: ec2.instances

And my /etc/ansible/hosts has a single entry:
localhost ansible_connection=local

I run this and it fails. Similar posts on this list suggest I have out of date modules, but I don’t know where these might be or how they might have got there?
langfords@langfords4$ ansible-playbook ansible-ec2.yaml -vvvv

PLAY [Create a sandbox instance] **********************************************

TASK: [Launch instance] *******************************************************
<127.0.0.1> EXEC [‘/bin/sh’, ‘-c’, ‘mkdir -p $HOME/.ansible/tmp/ansible-1381143657.34-222947650962172 && chmod a+rx $HOME/.ansible/tmp/ansible-1381143657.34-222947650962172 && echo $HOME/.ansible/tmp/ansible-1381143657.34-222947650962172’]
<127.0.0.1> REMOTE_MODULE ec2 keypair=langfords group=sg-1edd3571 instance_type=m1.small image=ami-c7c0d6b3 wait=true region=eu-west-1
<127.0.0.1> PUT /tmp/tmpocPJSL TO /home/langfords/.ansible/tmp/ansible-1381143657.34-222947650962172/ec2
<127.0.0.1> EXEC [‘/bin/sh’, ‘-c’, ‘/usr/bin/python /home/langfords/.ansible/tmp/ansible-1381143657.34-222947650962172/ec2; rm -rf /home/langfords/.ansible/tmp/ansible-1381143657.34-222947650962172/ >/dev/null 2>&1’]
failed: [localhost] => {“failed”: true, “parsed”: false}
invalid output was: Traceback (most recent call last):
File “/home/langfords/.ansible/tmp/ansible-1381143657.34-222947650962172/ec2”, line 1549, in
main()
File “/home/langfords/.ansible/tmp/ansible-1381143657.34-222947650962172/ec2”, line 612, in main
(instance_dict_array, new_instance_ids, changed) = create_instances(module, ec2)
File “/home/langfords/.ansible/tmp/ansible-1381143657.34-222947650962172/ec2”, line 437, in create_instances
res = ec2.run_instances(**params)
TypeError: run_instances() got an unexpected keyword argument ‘instance_profile_name’

FATAL: all hosts have already failed – aborting

PLAY RECAP ********************************************************************
to retry, use: --limit @/home/langfords/ansible-ec2.yaml.retry

localhost : ok=0 changed=0 unreachable=0 failed=1

2

How did you install ansible? The modules are (by default) installed in /usr/share/ansible, and it would appear that the cloud/ec2 file is out of date, as it’s erroring out on a new parameter that was added in 1.3.

3

Hi James,

I added the Ubuntu PPA mentioned on this page: http://www.ansibleworks.com/docs/intro_installation.html

And then just ran: sudo apt-get install ansible

Cheers
Simon

4

Hi James,

I checked the folder you mentioned and it seems to know about the parameter:

langfords@langfords4:/usr/share/ansible/cloud$ pwd
/usr/share/ansible/cloud
langfords@langfords4:/usr/share/ansible/cloud$ cat ec2 | grep instance_profile_name
instance_profile_name:
instance_profile_name = module.params.get(‘instance_profile_name’)
‘instance_profile_name’: instance_profile_name}
instance_profile_name = dict(),

5

So, I tried to force the module dir on the command line, and now it appears the issue has gone away, to be replaced by an auth issue:

langfords@langfords4:~/Dropbox/work/clouddeploy$ ansible-playbook -M /usr/share/ansible -i /etc/ansible/hosts ansible-ec2.yaml -vvvv

PLAY [Create a sandbox instance] **********************************************

TASK: [Launch instance] *******************************************************
<127.0.0.1> EXEC [‘/bin/sh’, ‘-c’, ‘mkdir -p $HOME/.ansible/tmp/ansible-1381222473.64-270653428553926 && chmod a+rx $HOME/.ansible/tmp/ansible-1381222473.64-270653428553926 && echo $HOME/.ansible/tmp/ansible-1381222473.64-270653428553926’]
<127.0.0.1> REMOTE_MODULE ec2 keypair=langfords group=sg-1edd3571 instance_type=m1.small image=ami-c7c0d6b3 wait=true region=eu-west-1
<127.0.0.1> PUT /tmp/tmp0BjAph TO /home/langfords/.ansible/tmp/ansible-1381222473.64-270653428553926/ec2
<127.0.0.1> EXEC [‘/bin/sh’, ‘-c’, ‘/usr/bin/python /home/langfords/.ansible/tmp/ansible-1381222473.64-270653428553926/ec2; rm -rf /home/langfords/.ansible/tmp/ansible-1381222473.64-270653428553926/ >/dev/null 2>&1’]
failed: [localhost] => {“failed”: true}
msg: No handler was ready to authenticate. 1 handlers were checked. [‘QuerySignatureV2AuthHandler’] Check your credentials

FATAL: all hosts have already failed – aborting

PLAY RECAP ********************************************************************
to retry, use: --limit @/home/langfords/ansible-ec2.yaml.retry

localhost : ok=0 changed=0 unreachable=0 failed=1

My /etc/ansible/hosts is still as below:
localhost ansible_connection=local

6

How have you configured your AWS credentials? Are you exporting them?

7

Ah, oops… I’d switched terminal windows… right, so back to the original prob.

I’ve run with -M to force the location and it makes no difference:

langfords@langfords4:~/Dropbox/work/clouddeploy$ ./run_ansible ansible-ec2.yaml -M /usr/share/ansible

PLAY [Create a sandbox instance] **********************************************

TASK: [Launch instance] *******************************************************
failed: [localhost] => {“failed”: true, “parsed”: false}
invalid output was: Traceback (most recent call last):
File “/home/langfords/.ansible/tmp/ansible-1381245971.29-257994512790004/ec2”, line 1549, in
main()
File “/home/langfords/.ansible/tmp/ansible-1381245971.29-257994512790004/ec2”, line 612, in main
(instance_dict_array, new_instance_ids, changed) = create_instances(module, ec2)
File “/home/langfords/.ansible/tmp/ansible-1381245971.29-257994512790004/ec2”, line 437, in create_instances
res = ec2.run_instances(**params)
TypeError: run_instances() got an unexpected keyword argument ‘instance_profile_name’

FATAL: all hosts have already failed – aborting

PLAY RECAP ********************************************************************
to retry, use: --limit @/home/langfords/ansible-ec2.yaml.retry

localhost : ok=0 changed=0 unreachable=0 failed=1

langfords@langfords4:~/Dropbox/work/clouddeploy$ cat /usr/share/ansible/cloud/ec2 | grep instance_profile_name
instance_profile_name:
instance_profile_name = module.params.get(‘instance_profile_name’)
‘instance_profile_name’: instance_profile_name}
instance_profile_name = dict(),

I got a hold of the 1.3.2 source and ran a diff between the two ec2 modules and get no diffs…

Is there anywhere else I can look to see what might be going on here? I don’t know python, so it’s hard for me to diagnose.

8

What version of boto do you have installed? This actually appears to be an issue where that parameter is being sent into the boto function run_instances(), but boto says it’s an invalid parameter name (it was introduced in boto 2.5.0).

9

Right. I have boto 2.2.2, frustrating that the ansible deb didn’t flag this. Seems to be non-trivial to get an ubuntu PPA for 12.04 which has >=2.5.0. Will find a way to get it on there and try again.

Thanks v much for you help.

Simon

10

Yup, that worked great. Installed with pip and now it runs. Except…

msg: InvalidGroup.NotFound: The security group ‘sg-1edd3571’ does not exist in default VPC ‘vpc-a86e25c3’

I’ve double checked and that security group is most definitely assigned to that VPC, which is the default VPC, which raises 2 questions:

  • Does ansible support VPC?

  • And because I’m about to stop using my default VPC, does it support a non-default VPC?

Thanks
Simon

11

The answer to both is yes, you should be able to specify the VPC with vpc_subnet_id. Do VPCs span regions in AWS? I’m a little bit ignorant of that detail myself, so it could be that the VPC you’ve defined isn’t in eu-west-1.

12

They don’t AFAIK, but it’s defo in eu-west-1. I hadn’t seen the vpc_subnet_id. When I went and checked the doco I realised i was putting a group_id into the group parameter. Sorted that out and it works.

Thanks for you help, I can delve a little further now.