How can you perform rolling updates in Kubernetes?

Performing rolling updates in Kubernetes is primarily achieved by updating the deployment configuration. When you modify the specifications of a Deployment, such as the container image version or other relevant fields, Kubernetes automatically manages the update process. It incrementally replaces the current Pods with new ones according to the updated configuration, ensuring that a specified number of Pods remain available during the transition.

This mechanism allows for a seamless update experience, minimizing downtime and maintaining application availability. Kubernetes ensures that the new Pods are ready and healthy before scaling down the old Pods, thus adhering to the principles of rolling updates, where the updates happen gradually and without disrupting service.

The other options do not represent the proper or efficient way to perform rolling updates in Kubernetes. For instance, using a special command for live reloading is not a standard practice in Kubernetes for managing updates. Manually terminating old Pods could lead to service disruption, as it removes existing Pods without ensuring that new instances are in place and running. Creating new Deployments for each version stops the old Pods and starts new ones, which does not align with the rolling update strategy and can create downtime. Thus, updating the deployment configuration stands out as the optimal and correct method to perform rolling updates in Kubernetes.