clearml-helm-charts/charts/clearml/templates/deployment-agent.yaml

{{- range $key, $value := .Values.agentGroups }}
{{- with $value }}
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: {{ include "clearml.fullname" $ }}-{{ .name }}-agent
  labels:
    {{- include "clearml.labels" $ | nindent 4 }}
spec:
  replicas: {{ .replicaCount }}
  selector:
    matchLabels:
      {{- include "clearml.selectorLabelsAgent" $ | nindent 6 }}
  template:
    metadata:
      {{- with .podAnnotations }}
      annotations:
        {{- toYaml . | nindent 8 }}
      {{- end }}
      labels:
        {{- include "clearml.selectorLabelsAgent" $ | nindent 8 }}
    spec:
      volumes:
        {{ if .clearmlConfig }}
        - name: agent-clearml-conf-volume
          secret:
            secretName: {{ .name }}-conf
            items:
            - key: clearml.conf
              path: clearml.conf
        {{ end }}
      initContainers:
        - name: init-agent-{{ .name }}
          image: "{{ .image.repository }}:{{ .image.tag | default $.Chart.AppVersion }}"
          command:
            - /bin/sh
            - -c
            - >
              set -x;
              while [ $(curl -sw '%{http_code}' "http://{{ include "clearml.fullname" $ }}-apiserver:{{ $.Values.apiserver.service.port }}/debug.ping" -o /dev/null) -ne 200 ] ; do
                echo "waiting for apiserver" ;
                sleep 5 ;
              done
      containers:
        - name: {{ $.Chart.Name }}-{{ .name }}
          image: "{{ .image.repository }}:{{ .image.tag }}"
          imagePullPolicy: {{ .image.pullPolicy }}
          securityContext:
            privileged: true
          resources:
            limits:
              nvidia.com/gpu:
                {{ .nvidiaGpusPerAgent }}
          env:
          - name: CLEARML_API_HOST
            value: 'http://{{ include "clearml.fullname" $ }}-apiserver:{{ $.Values.apiserver.service.port }}'
          - name: CLEARML_WEB_HOST
            value: 'http://{{ include "clearml.fullname" $ }}-webserver:{{ $.Values.webserver.service.port }}'
          - name: CLEARML_FILES_HOST
            value: 'http://{{ include "clearml.fullname" $ }}-fileserver:{{ $.Values.fileserver.service.port }}'
          - name: CLEARML_AGENT_GIT_USER
            value: {{ .clearmlGitUser}}
          - name: CLEARML_AGENT_GIT_PASS
            value: {{ .clearmlGitPassword}}
          - name: AWS_ACCESS_KEY_ID
            value: {{ .awsAccessKeyId}}
          - name: AWS_SECRET_ACCESS_KEY
            value: {{ .awsSecretAccessKey}}
          - name: AWS_DEFAULT_REGION
            value: {{ .awsDefaultRegion}}
          - name: AZURE_STORAGE_ACCOUNT
            value: {{ .azureStorageAccount}}
          - name: AZURE_STORAGE_KEY
            value: {{ .azureStorageKey}}
          - name: CLEARML_API_ACCESS_KEY
            valueFrom:
              secretKeyRef:
                name: clearml-conf
                key: tests_user_key
          - name: CLEARML_API_SECRET_KEY
            valueFrom:
              secretKeyRef:
                name: clearml-conf
                key: tests_user_secret
          command:
            - /bin/sh
            - -c
            - "apt-get update ;
              apt-get install -y curl python3-pip git;
              python3 -m pip install -U pip ;
              python3 -m pip install clearml-agent{{ .agentVersion}} ;
              CLEARML_AGENT_K8S_HOST_MOUNT=/root/.clearml:/root/.clearml clearml-agent daemon --queue {{ .queues}}"
          {{ if .clearmlConfig }}
          volumeMounts:
           - name: agent-clearml-conf-volume
             mountPath: /root/clearml.conf
             subPath: clearml.conf
             readOnly: true
          {{- end }}
      {{- with .nodeSelector }}
      nodeSelector:
        {{- toYaml . | nindent 8 }}
      {{- end }}
      {{- with .affinity }}
      affinity:
        {{- toYaml . | nindent 8 }}
      {{- end }}
      {{- with .tolerations }}
      tolerations:
        {{- toYaml . | nindent 8 }}
      {{- end }}
{{- end }}
{{- end }}