diff --git a/ray-cluster.yaml b/ray-cluster.yaml index f193910d59e10e8ad6b5d7298ce3367b99449925..287f20c39a8aa29431c18829c013b1b68e553785 100644 --- a/ray-cluster.yaml +++ b/ray-cluster.yaml @@ -1,11 +1,11 @@ -apiVersion: ray.io/v1alpha1 +apiVersion: ray.io/v1 kind: RayCluster metadata: name: raycluster-axol1tl namespace: axol1tl spec: - rayVersion: "2.0.0" - enableInTreeAutoscaling: true + rayVersion: "2.43.0" + enableInTreeAutoscaling: true # Allow autoscaling of workers headGroupSpec: serviceType: ClusterIP rayStartParams: @@ -15,7 +15,7 @@ spec: spec: containers: - name: ray-head - image: gitlab-registry.nrp-nautilus.io/mquinnan/axol1tl-hub:axol1tl-container + image: rayproject/ray:2.43.0 imagePullPolicy: Always resources: requests: @@ -31,12 +31,12 @@ spec: - name: axovol persistentVolumeClaim: claimName: axovol - tolerations: - - key: "nautilus.io/axol1tl" - operator: "Exists" - effect: "NoSchedule" + #tolerations: + #- key: "nautilus.io/axol1tl" + #operator: "Exists" + #effect: "NoSchedule" - workerGroupSpecs: + workerGroupSpecs: - groupName: worker-group rayStartParams: {} template: @@ -47,13 +47,13 @@ spec: imagePullPolicy: Always resources: requests: - cpu: "125" - memory: "512Gi" - nvidia.com/gpu: 8 + cpu: "8" + memory: "30Gi" + #nvidia.com/gpu: 8 # Request GPU limits: - cpu: "125" - memory: "512Gi" - nvidia.com/gpu: 8 + cpu: "8" + memory: "30Gi" + #nvidia.com/gpu: 8 volumeMounts: - name: axovol mountPath: /axovol @@ -61,9 +61,7 @@ spec: - name: axovol persistentVolumeClaim: claimName: axovol - tolerations: - - key: "nautilus.io/axol1tl" - operator: "Exists" - effect: "NoSchedule" - - + #tolerations: + #- key: "nautilus.io/axol1tl" + #operator: "Exists" + #effect: "NoSchedule"