provider/aws: Support suspending auto scaling group processes

[tpounds]

Adds support for suspending/resuming auto scaling group processes (e.g. Launch, Terminate, etc.).

see: http://docs.aws.amazon.com/AutoScaling/latest/DeveloperGuide/US_SuspendResume.html

[stack72]

Hi @tpounds,

I think this needs to change slightly, currently (in other resources), when we have an update command after a create command, the create should call the update command (e.g. resource_aws_instance.go)

func resourceAwsInstanceCreate(d *schema.ResourceData, meta interface{}) error {
    conn := meta.(*AWSClient).ec2conn

    instanceOpts, err := buildAwsInstanceOpts(d, meta)
    if err != nil {
        return err
    }
        ..........

    // Update if we need to
    return resourceAwsInstanceUpdate(d, meta)
}

func resourceAwsInstanceUpdate(d *schema.ResourceData, meta interface{}) error {
    conn := meta.(*AWSClient).ec2conn

}

I think will have to be the pattern in this case as well.

Thoughts?

Paul

[tpounds]

@stack72 I agree creating and then updating is cleaner. I'll ping you once I make the necessary changes to the resource.

[stack72]

@tpounds any luck with these updates yet? :)

[tpounds]

@stack72 Sorry I haven't provided an update yet. I took a look and updating the create/update code paths like you described requires consolidation of the auto scaling group wait for capacity logic (see: resource_aws_autoscaling_group.go:225, :380). In addition to that change, special care is required to prevent waiting when the Launch or Terminate policies are suspended since you might wait around forever for the capacity to spin up/down.

My high-level thoughts to fix this are as follows:

1. Consolidate wait for capacity logic (separate pull request)
2. Rebase this feature on config: bundle metadata requires global constraints #1 with consideration for Launch/Terminate policies as described above.

Thoughts?

[tpounds]

@stack72 I've created #5824 to cleanup the wait for capacity logic. I'll be rebasing this commit on that shortly.

[tpounds]

@stack72 Feature should be good to go now. I've rebased this change on #5824 so that should probably be merged first.

[sj26]

We've love support for this as well. Any reason this PR stagnated? Can I help get it over the line?

[stack72]

Closed via #10096

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.