How do you perform rolling updates and rollbacks in Kubernetes deployments?

Kubernetes deployments support rolling updates to update applications with zero downtime and rollback capabilities for quick recovery.\n\nRolling Update Strategy:\n• Gradually replaces old pods with new ones\n• Maintains application availability\n• Configurable update parameters\n• Automatic rollback on failure\n\nUpdate Configuration:\n• maxUnavailable: Maximum pods unavailable during update\n• maxSurge: Maximum additional pods during update\n• progressDeadlineSeconds: Timeout for update\n• revisionHistoryLimit: Number of old ReplicaSets to keep\n\nUpdate Commands:\nkubectl set image deployment/app container=image:v2\nkubectl rollout status deployment/app\nkubectl rollout history deployment/app\nkubectl rollout undo deployment/app --to-revision=1\n\nExample Strategy:\nspec:\n strategy:\n type: RollingUpdate\n rollingUpdate:\n maxUnavailable: 25%\n maxSurge: 25%\n progressDeadlineSeconds: 600\n revisionHistoryLimit: 10\n\nBest Practices:\n• Use health checks for update validation\n• Monitor application metrics during updates\n• Test rollback procedures\n• Implement automated rollback triggers\n• Use feature flags for safer deployments