diff --git a/.gitignore b/.gitignore index 380729e6..7dd7e039 100644 --- a/.gitignore +++ b/.gitignore @@ -150,4 +150,4 @@ cluster-config/cluster-config.yaml charts/graphistry-helm/templates/postgres/pg-sts.yaml -chart-bundle \ No newline at end of file +charts-aux-bundled \ No newline at end of file diff --git a/charts/graphistry-helm/templates/dask/dask-cluster.yml b/charts/graphistry-helm/templates/dask/dask-cluster.yml index a14c02c3..4eb7af49 100644 --- a/charts/graphistry-helm/templates/dask/dask-cluster.yml +++ b/charts/graphistry-helm/templates/dask/dask-cluster.yml @@ -29,17 +29,20 @@ spec: image: {{.Values.global.containerregistry.name}}/{{.Values.graphistry}}:{{.Values.daskcudaworker.repository}}-{{.Values.global.tag}}-{{.Values.cuda.version}} imagePullPolicy: {{.Values.global.imagePullPolicy }} args: - - dask-cuda-worker + - dask cuda worker - --interface - eth0 - --dashboard-address localhost:8787 - #- --resources GPU=10 - #- dask-scheduler:8786 + - --resources GPU=10,PROCESS=2 + - --nthreads 2 + - {{.Values.daskscheduler.location | quote }} command: - /entrypoints/rapids-entrypoint.sh env: - name: GRAPHISTRY_CPU_MODE value: {{.Values.graphistryCPUMode | quote }} + - name: SKIP_SERVER + value: "1" {{- range .Values.env }} - name: {{ .name }} value: {{ .value | quote }} diff --git a/charts/graphistry-helm/templates/dask/dask-cuda-worker-daemonset.yaml b/charts/graphistry-helm/templates/dask/dask-cuda-worker-daemonset.yaml index 8d3cd457..b919aa6b 100644 --- a/charts/graphistry-helm/templates/dask/dask-cuda-worker-daemonset.yaml +++ b/charts/graphistry-helm/templates/dask/dask-cuda-worker-daemonset.yaml @@ -41,17 +41,20 @@ spec: - "nexus" containers: - args: - - dask-cuda-worker + - dask cuda worker - --interface - eth0 - --dashboard-address localhost:8787 - #- --resources GPU=10 - - dask-scheduler:8786 + - --resources GPU=10,PROCESS=2 + - --nthreads 2 + - {{.Values.daskscheduler.location | quote }} command: - /entrypoints/rapids-entrypoint.sh env: - name: GRAPHISTRY_CPU_MODE value: {{.Values.graphistryCPUMode | quote }} + - name: SKIP_SERVER + value: "1" {{- range .Values.env }} - name: {{ .name }} value: {{ .value | quote }} diff --git a/charts/graphistry-helm/templates/forge-etl/forge-etl-python-daemonset.yaml b/charts/graphistry-helm/templates/forge-etl/forge-etl-python-daemonset.yaml index b8f60e0c..73978127 100644 --- a/charts/graphistry-helm/templates/forge-etl/forge-etl-python-daemonset.yaml +++ b/charts/graphistry-helm/templates/forge-etl/forge-etl-python-daemonset.yaml @@ -97,7 +97,7 @@ spec: command: - sh - -c - - "curl -f http://localhost:8080/cudfhealth && curl -f http://localhost:8080/blazinghealth" + - "curl -f http://localhost:8080/cudfhealth && curl -f http://localhost:8080/dasksqlhealth" failureThreshold: 3 initialDelaySeconds: 180 periodSeconds: 120 diff --git a/charts/graphistry-helm/templates/nginx/nginx-deployment.yaml b/charts/graphistry-helm/templates/nginx/nginx-deployment.yaml index f68e7fb1..d01d3848 100644 --- a/charts/graphistry-helm/templates/nginx/nginx-deployment.yaml +++ b/charts/graphistry-helm/templates/nginx/nginx-deployment.yaml @@ -207,7 +207,7 @@ spec: - name: PROMETHEUS_PORT value: "4040" {{- if eq .Values.global.containerregistry.name "docker.io/graphistry" }} - image: quay.io/martinhelmich/prometheus-nginxlog-exporter:v1.9.2 #DockerHub + image: ghcr.io/martin-helmich/prometheus-nginxlog-exporter/exporter:v1.9.2 #DockerHub {{ else }} image: {{.Values.global.containerregistry.name}}/prometheus-nginxlog-exporter:v1.9.2 {{- end }} diff --git a/charts/graphistry-helm/values.yaml b/charts/graphistry-helm/values.yaml index 0598bb6c..f5ad9a23 100644 --- a/charts/graphistry-helm/values.yaml +++ b/charts/graphistry-helm/values.yaml @@ -406,7 +406,7 @@ dask: #dask-scheduler repository name daskscheduler: repository: etl-server-python #dask-scheduler repository name - location: "dask-scheduler:8786" #dask-scheduler location + location: dask-scheduler:8786 #dask-scheduler location vgpu: false #enables vgpu mode for bigger than memory workloads on VGPU diff --git a/charts/postgres-cluster/templates/postgres-cluster.yaml b/charts/postgres-cluster/templates/postgres-cluster.yaml index f0de4ad2..8deb21be 100644 --- a/charts/postgres-cluster/templates/postgres-cluster.yaml +++ b/charts/postgres-cluster/templates/postgres-cluster.yaml @@ -48,11 +48,11 @@ spec: {{- if eq .Values.global.multiNode true }} accessModes: - "ReadWriteMany" - storageClassName: postgres-longhorn-{{ .Release.Namespace }} + storageClassName: postgres-longhorn {{- else }} accessModes: - "ReadWriteOnce" - storageClassName: retain-sc-{{ .Release.Namespace }} + storageClassName: retain-sc {{- end }} resources: requests: diff --git a/charts/values-overrides/internal/eks-dev-values.yaml b/charts/values-overrides/internal/eks-dev-values.yaml index d1257db4..0ee092ab 100644 --- a/charts/values-overrides/internal/eks-dev-values.yaml +++ b/charts/values-overrides/internal/eks-dev-values.yaml @@ -30,7 +30,7 @@ k8sDashboard: global: provisioner: ebs.csi.aws.com - tag: v2.40.3 + tag: v2.40.27 nodeSelector: {"accelerator": "nvidia"} #change log levels #logs: