When a group is created, an egress_rule ALLOW ALL to 0.0.0.0/0 is added
automatically but it's not reflected in the object returned by the AWS API
call. After creation we re-read the group for getting an updated object.
Suppose a pair of groups, A and B, depending on each other. One solution
for breaking the circular dependency at playbook level:
- declare group A without dependencies
- declare group B depending on A
- declare group A depending on B
This patch breaks the dependency at module level. Whenever a depended-on
group is missing it's first created. This approach requires only two tasks:
- declare group A depending on B (group B will be auto created)
- declare group B depending on A
When creating a group EC2 requires you to pass the group description. In
order to fullfil this, rules now accept the `group_desc` param. Note
that group description can't be changed once the group is created so
it's nice to keep descriptions in sync.
Concrete example:
- ec2_group:
name: mysql-client
description: MySQL Client
rules_egress:
- proto: tcp
from_port: 3306
to_port: 3306
group_name: mysql-server
group_desc: MySQL Server
- ec2_group:
name: mysql-server
description: MySQL Server
rules:
- proto: tcp
from_port: 3306
to_port: 3306
group_name: mysql-client
* Added desired_capacity and vpc_zone_identifier to ec2_asg
* Use ec2_argument_spec() method and then remove unnecessary
declarations from argument_spec
* Remove AWS_REGIONS declaration
* Rename block_device_mappings to volumes to be consistent with ec2
* Remove all pep8 warnings except line length and continuation indent
* Use updated module_utils/ec2.py to add profile and security_token
support
* Remove mandatory arguments for delete to make launchconfig deletion
work
* Handle existing launch configurations better
* Improve output information
* Improve documentation
Had to shoot the recently merged nova_group module in the head temporarily as it contained a dict comprehension, which means it can't work on all the platforms
and was also breaking docs builds on CentOS. Will engage with list about that shortly.
The new present state just makes sure that a container exists, not that
it's running, although it get started one creation.
This is very useful for data volumes. This also changes the old
present, now running (default) state to only create the container if
it's not found, otherwise it just get started.
See also discussion on mailinglist:
https://groups.google.com/forum/#!topic/ansible-devel/jB84gdhPzLQ
This closes#6395
