The L2 node shall use the newly exposed compute_step_many as defined in #3800 (closed).
compute_step_many