Improve memory and cpu performance of containers

[mhl787156]

This issue is a catch all for general container performance improvements. current improvements include:

• Multi-staged builds to reduce memory footprint, especially for the larger containers.
  • Includes removing unused files for runtime (e.g. build folders and gazebo models)
• Anything to reduce traffic and hardware requirements. Running a SITL/Gazebo/mavros sim with foxglove and gazebo web and the controller can get very heavy on any machine causing performance issues, in particular network bottleneck ros2 performance issues (parameter and node discovery).

Feel free to add more...

[mhl787156]

In particular with gazebo, it may be worth looking into making it an nvidia GPU container. See these links for details:

• https://github.com/sunsided/ros-gazebo-gpu-docker
• https://docs.aws.amazon.com/robomaker/latest/dg/running-gpu-sample-application-ros2-gazebo.html

[mhl787156]

Memory addressed with #161 Still need to address computation requirements.

Perhaps looking into Docker run or build options.