You are tasked with deploying a critical application to a Google Cloud Platform (GCP) environment, specifically on a Compute Engine virtual machine (VM) within a managed instance group. The application requires high availability, implying it must remain operational at all times. However, it is crucial that only one instance of this VM is running within the entire GCP project to avoid conflicts and resource wastage. How should you set up the managed instance group to meet these requirements?