Awslogs agent stop fails

[mbloch1986]

Describe the bug
The AMI baking fails sometimes because it can not stop the Awslogs agent at the end of the AMI baking.

At the end of the AMI baking we are starting the awslogs agent and stopping it instantly. We should analyse if the start of the awslogs agent is needed if not why starting it in the first place ?

�[0;32m    aws: �[mNotice: /Stage[main]/Main/Exec[Stop Cloudwatchlogs agent]/returns: Job for awslogs.service canceled.�[0m�[0m
�[0;32m    aws: �[1;31mError: 'systemctl stop awslogs' returned 1 instead of one of [0]�[0m�[0m
�[0;32m    aws: �[1;31mError: /Stage[main]/Main/Exec[Stop Cloudwatchlogs agent]/returns: change from 'notrun' to ['0'] failed: 'systemctl stop awslogs' returned 1 instead of one of [0]�[0m�[0m

To Reproduce
Steps to reproduce the behavior:
Start an AMI baking

Expected behavior
AMI baking should not fail because it couldn't stop the awslogsd service

Screenshots

Environment (please complete the following information if relevant):

• OS RHEL7, amz2
• Packer-AEM : any versoin

Additional context
Add any other context about the problem here.