publish: true gemini: true tags:

• kubernetes
• site_reliability created: ‘2026-05-04 14:40’ last_modified: ‘2026-05-04 14:40:00’ status: evergreen confidence: fact

Rolling migrations prevent homelab downtime

By cordoning and draining a single node, shifting workloads, and upgrading the base OS, a cluster maintains quorum and application availability. This methodical SRE practice ensures that High-quality infrastructure is invisible when working correctly.¹

Footnotes

1. Kubernetes Documentation. Safely Draining a Node. ↩