Skip to content

Conversation

@Manan-Kothari
Copy link

Fixes #63

For example if you now define the topologySpreadConstraint like so

worker:
  topologySpreadConstraints:
    - maxSkew: 1
      topologyKey: topology.kubernetes.io/zone
      whenUnsatisfiable: DoNotSchedule
      labelSelector:
        matchLabels:
          service: ray-worker

this outputs the correct manifest with the TSC.

  workerGroupSpecs:
  - rayStartParams:
      {}
    replicas: 1
    minReplicas: 1
    maxReplicas: 3
    numOfHosts: 1
    groupName: workergroup
    template:
      spec:
        imagePullSecrets:
          []
        containers:
          - 
            volumeMounts:
            - mountPath: /tmp/ray
              name: log-volume
            name: ray-worker
            image: rayproject/ray:2.41.0
            imagePullPolicy: IfNotPresent
            resources:
              limits:
                cpu: "1"
                memory: 1G
              requests:
                cpu: "1"
                memory: 1G
            securityContext:
              {}
        
        volumes:
          - emptyDir: {}
            name: log-volume
        topologySpreadConstraints:
          - labelSelector:
              matchLabels:
                service: ray-worker
            maxSkew: 1
            topologyKey: topology.kubernetes.io/zone
            whenUnsatisfiable: DoNotSchedule

Also fixes TSC for head node and additional worker groups as well.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Bug: TopologySpreadConstraints are currently broken due to being defined as an empty map

1 participant