Your Dataflow pipeline, situated in europe-west4, ingests messages from a Pub/Sub topic and stores the results in a BigQuery dataset located in the EU. The pipeline utilizes a maximum of 3 workers, each with an n1-standard-1 instance type. During peak periods, you observe that the pipeline struggles to process records promptly, with all 3 workers operating at maximum CPU utilization. What are two actions you can take to enhance the performance of your pipeline? (Select two.)