Fix default cleanup state machine

[ghost]

While looking to refactor CheckPoweredOn state, I realized that the cleanup state machine was targeting the destination VM for power on, while it should target source VM. The logic is similar to what is implemented in poweron.rb in #378. This PR implements it in checkpoweredon.rb.

Doing so makes the checkpoweredon.rb method generic, so it's moved to /Transformation/Infrastructure/VM/Common class. This requires changing two state machines: PostTransform and TransformationCleanup.

Changing TransformationCleanup state machine removes the need for destination_ems_type state var. The method assesstransformationcleanup.rb only purpose was to set this state var, so it can be removed.

When running tests, I also realized that service_template_transformation_plan_task_id attribute was missing in $evm.root. Somehow, the change is weightedupdatestatus.rb to add it to the cleanup request was not pushed in #378, so I'm fixing it here.

Associated RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=1599997

[ghost]

@miq-bot add-label transformation
@miq-bot add-label bug
@miq-bot add-label gaprindashvili/yes
@miq-bot add-label blocker

[miq-bot]

Checked commit fabiendupont@baa34b4 with ruby 2.3.3, rubocop 0.52.1, haml-lint 0.20.0, and yamllint 1.10.0
2 files checked, 1 offense detected

content/automate/ManageIQ/Transformation/Infrastructure/VM/Common.class/methods/checkpoweredon.rb

• ❗ - Line 27, Col 15 - Style/RescueStandardError - Avoid rescuing without specifying an error class.

[gmcculloug]

We should break this down into methods which would make the logic easier to read.

Move this first if-else block into a method that returns both task and vm:

def task_and_vm
  if @handle.root['service_template_transformation_plan_task'].blank?
    task = @handle.vmdb(:service_template_transformation_plan_task).find_by(:id => @handle.root['service_template_transformation_plan_task_id'])
    vm = task.source if task.present?
  else
    task = @handle.root['service_template_transformation_plan_task']
    vm = @handle.vmdb(:vm).find_by(:id => task.get_option(:destination_vm_id)) if task.present?
  end

  return task, vm
end

Then the first part of main becomes:

task, vm = task_and_vm

If you think this is too much change it could be in a followup refactoring, but we need to move away for having large single method scripts in general.

[ghost]

The code is identical in checkpoweredon.rb and poweron.rb. I'd propose moving task_and_vm to a embedded utility method. It's a bigger change that may require another PR. I'd go for a followup refactoring.

[gmcculloug]

For readability add spaces between lines when the context changes:

I would think something like this:

def main
  task, vm = task_and_vm
  return if vm.blank?

  @handle.log(:info, "Target VM: #{vm.name} [#{vm.vendor}]")

  return if task.get_option(:source_vm_power_state) != 'on'

  if vm.power_state != 'on'
    @handle.root["ae_result"] = "retry"
    @handle.root["ae_retry_interval"] = "15.seconds"
  end
rescue => e
  @handle.set_state_var(:ae_state_progress, 'message' => e.message)
  raise
end

[simaishi]

Gaprindashvili backport details:

$ git log -1
commit 3d43a639999bb0194e34c8a80a7d6f360dfaa32e
Author: Greg McCullough <gmccullo@redhat.com>
Date:   Tue Jul 31 20:48:51 2018 -0400

    Merge pull request #382 from fdupont-redhat/v2v_fix_cleanup_state_machine_poweron
    
    Fix default cleanup state machine
    (cherry picked from commit 3e6a3f656cac57ad1b9ec2c1a6d1e8cdde44e6f4)
    
    https://bugzilla.redhat.com/show_bug.cgi?id=1608758