A client wants to deploy a MongoDB database in Google Cloud for a high-traffic web application. They want to optimize for high availability and performance. They also have a continuous, substantial stream of incoming writes. What is the most suitable strategy for this scenario?