Problem Statement
After deploying a new version of your micro-services platform on Kubernetes the cost of cloud resources spiked unexpectedly. What immediate investigations and corrective actions do you take to control cost and prevent similar episodes?
Explanation
First I’d identify which resources grew unexpectedly: check node count, instance types, pod replica counts, auto-scaler behaviour, storage usage and provider billing alerts. Then I’d correlate with deployment time and application logs to see if scaling logic mis-behaved or replica count increased due to unhealthy pods being restarted. I’d temporarily scale down non-critical workloads, pause auto-scaling if needed, adjust resource requests/limits to match actual usage, and enable cost-monitoring alerts. I’d also review CI/CD pipelines for mis-configuration (e.g., replicas set to high defaults). Long-term I’d implement cost-governance policies, tag resources, use right-sizing tools and review auto-scaler settings. This type of operational scenario is commonly asked for senior devops roles. :contentReference[oaicite:5]{index=5}
