Year-End Kubernetes Cost Cleanup Checklist

Close the year with a tidy cluster and a clear budget. This checklist takes 90 minutes and prevents surprises in January.

1) Prune and archive

• Delete unused namespaces, preview environments, and orphaned PVCs.
• Reduce log/metric retention to sane defaults; archive cold data to object storage.
• Remove dormant feature flags and cron jobs that still run nightly.

2) Refresh guardrails

• Reapply ResourceQuota and LimitRange templates; tighten staging more than prod.
• Validate owner/team labels across workloads; auto-fix or block missing metadata.
• Audit NetworkPolicy and PodSecurity to ensure new services are covered.

3) Rebalance capacity

• Rightsize node groups based on current p95 utilization; drop oversized instances.
• Reevaluate spot vs on-demand mix for each pool; diversify instance types.
• Clean up unused load balancers, IPs, and DNS records.

4) Budget and plan

• Set monthly budgets per product with ±10% guardrails.
• Schedule quarterly cost reviews and cost game days.
• Update commitment plans (reserved/spot) based on steady workloads.

5) Communicate

• Share before/after spend and waste reductions.
• Publish the Q1 optimization backlog with owners and projected savings.
• Thank teams that moved the cost needle—make it a cultural win.

A short cleanup now saves weeks of firefighting later. Start the year with lean defaults, fresh guardrails, and a clear plan.***