Day-15 - Kubernetes scheduling — Node Affinity & Anti-Affinity

 Day-15 - Kubernetes scheduling — Node Affinity & Anti-Affinity

#Day-15

Today I explored another key part of Kubernetes scheduling — Node Affinity & Anti-Affinity ⚙️
These mechanisms go beyond Taints & Tolerations and give us intelligent control over where and how pods are scheduled across nodes.

Key Takeaways:
1️⃣ Node Affinity — allows you to target specific nodes based on labels (e.g., disk=ssd, zone=us-east-1a).
2️⃣ Required vs Preferred — requiredDuringSchedulingIgnoredDuringExecution enforces strict placement, while preferredDuringSchedulingIgnoredDuringExecution defines flexible preferences.
3️⃣ Node Anti-Affinity — helps you avoid certain nodes (e.g., don’t schedule in zone=us-east-1a).
4️⃣ Pod Anti-Affinity — ensures similar pods (like replicas) don’t run on the same node — improving HA and resilience.
5️⃣ Combine these with Taints & Tolerations for fine-grained, intelligent scheduling policies.