As a DevOps Engineer, you're optimizing a web application that often faces sudden traffic spikes. The app uses multiple microservices, and efficient resource use is crucial. Which strategy best ensures optimal performance and resource efficiency during these spikes?