Your current application, which is deployed on Google Kubernetes Engine (GKE), utilizes multiple pods distributed across four GKE n1-standard-2 nodes. You are required to deploy more pods that specifically require n2-highmem-16 nodes, and you must do this without causing any downtime. What actions should you take to achieve this?