Book notes

Appearance

Affinity ​

To destribute the pods across different availability zones, we can use the nodeAffinity property.

Availability zones are a way to group resources in a data center.

They are usually used to group resources that are close to each other and have a low latency between them.

Distributing pods across availability zones is a way to make sure that if one availability zone goes down, the application will still be available in the other availability zones.

Affinity section in the pod definition:

yaml
...
spec:
  affinity:
    podAntiAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        - labelSelector:
            matchExpressions:
              - key: app
                operator: In
                values:
                  - my-app
          topologyKey: kubernetes.io/hostname
...
...
spec:
  affinity:
    podAntiAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        - labelSelector:
            matchExpressions:
              - key: app
                operator: In
                values:
                  - my-app
          topologyKey: kubernetes.io/hostname
...

The podAntiAffinity property is used to make sure that pods with the same label are not scheduled on the same node.

The requiredDuringSchedulingIgnoredDuringExecution property is used to make sure that the pods are not scheduled on the same node.

Preffered vs Required ​

Preffered means that the scheduler will try to schedule the pods on different nodes, but if it is not possible, it will schedule them on the same node.

Required means if it can't find a suitable node, it will not schedule the pod at all.

Node Affinity ​

Node affinity is similar to pod affinity, but it is used to schedule pods based on the node labels.

yaml
...
spec:
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
          - matchExpressions:
              - key: disktype
                operator: In
                values:
                  - ssd
...
...
spec:
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
          - matchExpressions:
              - key: disktype
                operator: In
                values:
                  - ssd
...

Inter-pod Affinity ​

Inter-pod affinity is used to schedule pods based on the labels of other pods.

Weighted Affinity ​

If scheduler found multiple nodes that match the affinity rules, it will use the weight property to decide which node to use.

Nodes with higher weight will be preferred.

Resources ​