gym-gazebo how does agent react to action_id?

edit

Hi does someone mind please explaining how the Gazebo environment reacts to the "step(action_id)" function?

Specifically, when testing the trained algorithm on a turtlebot, does the turtlebot receive commands that change extremely frequently because its constantly sensing and acting, or does the turtlebot wait for sometime before implementing the next action?

Thank you for your help in advance!