k8s: spread pods using podAntiAffinity vs topologySpreadConstraints

I used to force gitlab-runner to different node using podAntiAffinity

kind: Deployment
metadata:
  name: gitlab-runner
spec:
  ...
  template:
    ...
    spec:
      affinity:
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 100
            podAffinityTerm:
              labelSelector:
                matchExpressions:
                - key: "app"
                  operator: "In"
                  values:
                  - "gitlab-runner"
              topologyKey: "kubernetes.io/hostname"

This forces each gitlab-runner pod to different node

Using topologySpreadConstraints , I can spread pods among nodes. e.g.

• node 1: 3 pods running
• node 2: 4 pods running

apiVersion: apps/v1
kind: Deployment
metadata:
  name: gitlab-runner
spec:
  ...
  template:
    ...
    spec:
      topologySpreadConstraints:
        - maxSkew: 1
          topologyKey: kubernetes.io/hostname
          whenUnsatisfiable: DoNotSchedule
          labelSelector:
            matchLabels:
              app: gitlab-runner