Improvements k8sagent (#54)

* Fixed: image versions

* Changed: chart version

* Changed: readme update by helm-docs

.github/helm-docs.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
 CHART_DIRS="$(git diff --find-renames --name-only "$(git rev-parse --abbrev-ref HEAD)" remotes/origin/main -- 'charts' | grep '[cC]hart.yaml' | sed -e 's#/[Cc]hart.yaml##g')"
-HELM_DOCS_VERSION="1.5.0"
+HELM_DOCS_VERSION="1.7.0"
 
 curl --silent --show-error --fail --location --output /tmp/helm-docs.tar.gz https://github.com/norwoodj/helm-docs/releases/download/v"${HELM_DOCS_VERSION}"/helm-docs_"${HELM_DOCS_VERSION}"_Linux_x86_64.tar.gz
 tar -xf /tmp/helm-docs.tar.gz helm-docs

charts/clearml/Chart.lock

@@ -7,6 +7,6 @@ dependencies:
   version: 10.3.4
 - name: elasticsearch
   repository: https://helm.elastic.co
-  version: 7.10.1
-digest: sha256:aefd3992b2ab085161e4cca35c6f73dd33f8d19272a9405b5ee4e8c2a0e79bba
-generated: "2021-01-05T14:26:33.629164+01:00"
+  version: 7.16.2
+digest: sha256:ac733cb02d50e8398c1d2832988333896f1c7b765c19a0f1eea5e9b24bdb8207
+generated: "2022-01-05T07:52:34.913745+01:00"

charts/clearml/Chart.yaml

@@ -2,8 +2,8 @@ apiVersion: v2
 name: clearml
 description: MLOps platform
 type: application
-version: "3.2.2"
-appVersion: "1.1.1"
+version: "3.6.0"
+appVersion: "1.2.0"
 home: https://clear.ml
 icon: https://raw.githubusercontent.com/allegroai/clearml/master/docs/clearml-logo.svg
 sources:
@@ -18,14 +18,14 @@ keywords:
   - mlops
 dependencies:
   - name: redis
-    version: "~10.9.0"
+    version: "10.9.0"
     repository: "https://charts.bitnami.com/bitnami"
     condition: redis.enabled
   - name: mongodb
-    version: "~10.3.2"
+    version: "10.3.4"
     repository: "https://charts.bitnami.com/bitnami"
     condition: mongodb.enabled
   - name: elasticsearch
-    version: "~7.10.1"
+    version: "7.16.2"
     repository: "https://helm.elastic.co"
     condition: elasticsearch.enabled

charts/clearml/README.md

@@ -1,6 +1,6 @@
 # ClearML Ecosystem for Kubernetes
 
-![Version: 3.2.2](https://img.shields.io/badge/Version-3.2.2-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: 1.1.1](https://img.shields.io/badge/AppVersion-1.1.1-informational?style=flat-square)
+![Version: 3.6.0](https://img.shields.io/badge/Version-3.6.0-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: 1.2.0](https://img.shields.io/badge/AppVersion-1.2.0-informational?style=flat-square)
 
 MLOps platform
 
@@ -107,9 +107,9 @@ For detailed instructions, see the [Optional Configuration](https://github.com/a
 
 | Repository | Name | Version |
 |------------|------|---------|
-| https://charts.bitnami.com/bitnami | mongodb | ~10.3.2 |
-| https://charts.bitnami.com/bitnami | redis | ~10.9.0 |
-| https://helm.elastic.co | elasticsearch | ~7.10.1 |
+| https://charts.bitnami.com/bitnami | mongodb | 10.3.4 |
+| https://charts.bitnami.com/bitnami | redis | 10.9.0 |
+| https://helm.elastic.co | elasticsearch | 7.16.2 |
 
 ## Values
 
@@ -163,6 +163,19 @@ For detailed instructions, see the [Optional Configuration](https://github.com/a
 | agentGroups.agent-group-gpu.replicaCount | int | `0` |  |
 | agentGroups.agent-group-gpu.tolerations | list | `[]` |  |
 | agentGroups.agent-group-gpu.updateStrategy | string | `"Recreate"` |  |
+| agentk8sglue.defaultDockerImage | string | `"nvidia/cuda:11.3.1-cudnn8-runtime-ubuntu20.04"` |  |
+| agentk8sglue.enabled | bool | `false` |  |
+| agentk8sglue.id | string | `"k8s-agent"` |  |
+| agentk8sglue.image.repository | string | `"allegroai/clearml-agent-k8s"` |  |
+| agentk8sglue.image.tag | string | `"aws-latest-1.21"` |  |
+| agentk8sglue.maxPods | int | `10` |  |
+| agentk8sglue.podTemplate.env | list | `[]` |  |
+| agentk8sglue.podTemplate.nodeSelector | object | `{}` |  |
+| agentk8sglue.podTemplate.resources | object | `{}` |  |
+| agentk8sglue.podTemplate.tolerations | list | `[]` |  |
+| agentk8sglue.podTemplate.volumes | list | `[]` |  |
+| agentk8sglue.queue | string | `"aws-instances"` |  |
+| agentk8sglue.serviceAccountName | string | `"default"` |  |
 | agentservices.affinity | object | `{}` |  |
 | agentservices.agentVersion | string | `""` |  |
 | agentservices.awsAccessKeyId | string | `nil` |  |
@@ -196,7 +209,7 @@ For detailed instructions, see the [Optional Configuration](https://github.com/a
 | apiserver.extraEnvs | list | `[]` |  |
 | apiserver.image.pullPolicy | string | `"IfNotPresent"` |  |
 | apiserver.image.repository | string | `"allegroai/clearml"` |  |
-| apiserver.image.tag | string | `"1.1.1"` |  |
+| apiserver.image.tag | string | `"1.2.0"` |  |
 | apiserver.livenessDelay | int | `60` |  |
 | apiserver.nodeSelector | object | `{}` |  |
 | apiserver.podAnnotations | object | `{}` |  |
@@ -256,7 +269,7 @@ For detailed instructions, see the [Optional Configuration](https://github.com/a
 | fileserver.extraEnvs | list | `[]` |  |
 | fileserver.image.pullPolicy | string | `"IfNotPresent"` |  |
 | fileserver.image.repository | string | `"allegroai/clearml"` |  |
-| fileserver.image.tag | string | `"1.1.1"` |  |
+| fileserver.image.tag | string | `"1.2.0"` |  |
 | fileserver.nodeSelector | object | `{}` |  |
 | fileserver.podAnnotations | object | `{}` |  |
 | fileserver.replicaCount | int | `1` |  |
@@ -268,11 +281,14 @@ For detailed instructions, see the [Optional Configuration](https://github.com/a
 | fileserver.storage.data.size | string | `"50Gi"` |  |
 | fileserver.tolerations | list | `[]` |  |
 | ingress.annotations | object | `{}` |  |
+| ingress.api.annotations | object | `{}` |  |
 | ingress.api.hostName | string | `"api.clearml.127-0-0-1.nip.io"` |  |
 | ingress.api.tlsSecretName | string | `""` |  |
+| ingress.app.annotations | object | `{}` |  |
 | ingress.app.hostName | string | `"app.clearml.127-0-0-1.nip.io"` |  |
 | ingress.app.tlsSecretName | string | `""` |  |
 | ingress.enabled | bool | `false` |  |
+| ingress.files.annotations | object | `{}` |  |
 | ingress.files.hostName | string | `"files.clearml.127-0-0-1.nip.io"` |  |
 | ingress.files.tlsSecretName | string | `""` |  |
 | ingress.name | string | `"clearml-server-ingress"` |  |
@@ -306,7 +322,7 @@ For detailed instructions, see the [Optional Configuration](https://github.com/a
 | webserver.extraEnvs | list | `[]` |  |
 | webserver.image.pullPolicy | string | `"IfNotPresent"` |  |
 | webserver.image.repository | string | `"allegroai/clearml"` |  |
-| webserver.image.tag | string | `"1.1.1"` |  |
+| webserver.image.tag | string | `"1.2.0"` |  |
 | webserver.nodeSelector | object | `{}` |  |
 | webserver.podAnnotations | object | `{}` |  |
 | webserver.replicaCount | int | `1` |  |

charts/clearml/ci/README.md

@@ -0,0 +1,7 @@
+Place values files with different values in this directory to ensure these cases are tested by the CI as well.
+
+https://github.com/helm/chart-testing/blob/main/doc/ct_install.md
+
+```
+"Charts may have multiple custom values files matching the glob pattern '*-values.yaml' in a directory named 'ci' in the root of the chart's directory. The chart is installed and tested for each of these files. If no custom values file is present, the chart is installed and tested with defaults."
+```

charts/clearml/ci/default-values.yaml

@@ -0,0 +1 @@
+# empty so default values.yaml gets tested

charts/clearml/ci/k8sagent-values.yaml

@@ -0,0 +1,2 @@
+agentk8sglue:
+  enabled: true

charts/clearml/templates/configmap-agentk8s-template.yaml

@@ -0,0 +1,52 @@
+{{- if .Values.agentk8sglue.enabled }}
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: k8sagent-pod-template
+data:
+  template.yaml: |
+    apiVersion: v1
+    metadata:
+      namespace: {{ .Release.namespace }}
+    spec:
+      serviceAccountName: {{ .Values.agentk8sglue.serviceAccountName }}
+      volumes:
+      {{- range .Values.agentk8sglue.podTemplate.volumes }}
+      - name: {{ .name }}
+        persistentVolumeClaim:
+          claimName: {{ .name }}
+      {{- end }} 
+      containers:
+      - resources:
+          {{- toYaml .Values.agentk8sglue.podTemplate.resources | nindent 10 }}
+        ports:
+        - containerPort: 10022
+        volumeMounts: 
+        {{- range .Values.agentk8sglue.podTemplate.volumes }}
+        - mountPath: {{ .path }}
+          name: {{ .name }}
+        {{- end }} 
+        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"
+        - name: CLEARML_FILES_HOST
+          value: "http://{{ include "clearml.fullname" . }}-fileserver:{{ .Values.fileserver.service.port }}"
+        - name: CLEARML_API_ACCESS_KEY
+          valueFrom:
+            secretKeyRef:
+              name: clearml-conf
+              key: apiserver_key
+        - name: CLEARML_API_SECRET_KEY
+          valueFrom:
+            secretKeyRef:
+              name: clearml-conf
+              key: apiserver_secret
+        {{- toYaml .Values.agentk8sglue.podTemplate.env | nindent 8 }}
+      tolerations:
+        {{- toYaml .Values.agentk8sglue.podTemplate.tolerations | nindent 8 }}
+      nodeSelector:
+        {{- toYaml .Values.agentk8sglue.podTemplate.nodeSelector | nindent 8 }}
+{{- end }}
+

charts/clearml/templates/deployment-agentk8s.yaml

@@ -0,0 +1,62 @@
+{{- if .Values.agentk8sglue.enabled }}
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: "{{ include "clearml.fullname" . }}-k8sagent"
+  labels:
+    app: k8sagent
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: k8sagent
+  template:
+    metadata:
+      labels:
+        app: k8sagent
+    spec:
+      containers:
+      - name: k8s-glue-container
+        image: "{{ .Values.agentk8sglue.image.repository }}:{{ .Values.agentk8sglue.image.tag }}"
+        imagePullPolicy: Always
+        command: ["/bin/bash", "-c", "export PATH=$PATH:$HOME/bin; source /root/.bashrc && /root/entrypoint.sh"]
+        volumeMounts:
+          - name: k8sagent-pod-template
+            mountPath: /root/template
+        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"
+          - name: CLEARML_FILES_HOST
+            value: "http://{{ include "clearml.fullname" . }}-fileserver:{{ .Values.fileserver.service.port }}"
+          - name: K8S_GLUE_MAX_PODS
+            value: "{{.Values.agentk8sglue.maxPods}}"
+          - name: K8S_GLUE_QUEUE
+            value: "{{.Values.agentk8sglue.queue}}"
+          - name: K8S_GLUE_EXTRA_ARGS
+            value: "--template-yaml /root/template/template.yaml"
+          - name: CLEARML_API_ACCESS_KEY
+            valueFrom:
+              secretKeyRef:
+                name: clearml-conf
+                key: apiserver_key
+          - name: CLEARML_API_SECRET_KEY
+            valueFrom:
+              secretKeyRef:
+                name: clearml-conf
+                key: apiserver_secret
+          - name: CLEARML_WORKER_ID
+            value: "{{.Values.agentk8sglue.id}}"
+          - name: CLEARML_AGENT_UPDATE_REPO
+            value: ""
+          - name: FORCE_CLEARML_AGENT_REPO
+            value: ""
+          - name: CLEARML_DOCKER_IMAGE
+            value: "{{.Values.agentk8sglue.defaultDockerImage}}"
+      volumes:
+        - name: k8sagent-pod-template
+          configMap: 
+            name: k8sagent-pod-template
+{{- end }}
+

charts/clearml/templates/ingress-api.yaml

@@ -11,10 +11,13 @@ metadata:
   name: {{ include "clearml.fullname" . }}-api
   labels:
     {{- include "clearml.labels" . | nindent 4 }}
-  {{- with .Values.ingress.annotations }}
-  annotations:
-    {{- toYaml . | nindent 4 }}
+  {{- $annotations := .Values.ingress.annotations }}
+  {{- if .Values.ingress.api.annotations }}
+  {{- $annotations = mergeOverwrite $annotations .Values.ingress.api.annotations }}
   {{- end }}
+  annotations:
+    {{- toYaml $annotations | nindent 4 }}
+
 spec:
   {{- if .Values.ingress.api.tlsSecretName }}
   tls:

charts/clearml/templates/ingress-app.yaml

@@ -11,10 +11,12 @@ metadata:
   name: {{ include "clearml.fullname" . }}-app
   labels:
     {{- include "clearml.labels" . | nindent 4 }}
-  {{- with .Values.ingress.annotations }}
-  annotations:
-    {{- toYaml . | nindent 4 }}
+  {{- $annotations := .Values.ingress.annotations }}
+  {{- if .Values.ingress.app.annotations }}
+  {{- $annotations = mergeOverwrite $annotations .Values.ingress.app.annotations }}
   {{- end }}
+  annotations:
+    {{- toYaml $annotations | nindent 4 }}
 spec:
   {{- if .Values.ingress.app.tlsSecretName }}
   tls:

charts/clearml/templates/ingress-files.yaml

@@ -11,10 +11,12 @@ metadata:
   name: {{ include "clearml.fullname" . }}-files
   labels:
     {{- include "clearml.labels" . | nindent 4 }}
-  {{- with .Values.ingress.annotations }}
-  annotations:
-    {{- toYaml . | nindent 4 }}
+  {{- $annotations := .Values.ingress.annotations }}
+  {{- if .Values.ingress.files.annotations }}
+  {{- $annotations = mergeOverwrite $annotations .Values.ingress.files.annotations }}
   {{- end }}
+  annotations:
+    {{- toYaml $annotations | nindent 4 }}
 spec:
   {{- if .Values.ingress.files.tlsSecretName }}
   tls:

charts/clearml/templates/rbac-agentk8s.yaml

@@ -0,0 +1,25 @@
+{{- if .Values.agentk8sglue.enabled }}
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: k8sagent-pods-access
+rules:
+  - apiGroups:
+      - ""
+    resources:
+      - pods
+    verbs: ["get", "list", "watch", "create", "patch", "delete"]
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  name: k8sagent-pods-access
+subjects:
+  - kind: ServiceAccount
+    name: default
+    namespace: {{ .Release.namespace }}
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: k8sagent-pods-access
+{{- end }}

charts/clearml/values.yaml

@@ -7,12 +7,15 @@ ingress:
   app:
     hostName: "app.clearml.127-0-0-1.nip.io"
     tlsSecretName: ""
+    annotations: {}
   api:
     hostName: "api.clearml.127-0-0-1.nip.io"
     tlsSecretName: ""
+    annotations: {}
   files:
     hostName: "files.clearml.127-0-0-1.nip.io"
     tlsSecretName: ""
+    annotations: {}
 
 secret:
   # -- Set for http_session field
@@ -56,7 +59,7 @@ apiserver:
   image:
     repository: "allegroai/clearml"
     pullPolicy: IfNotPresent
-    tag: "1.1.1"
+    tag: "1.2.0"
 
   extraEnvs: []
 
@@ -105,7 +108,7 @@ fileserver:
   image:
     repository: "allegroai/clearml"
     pullPolicy: IfNotPresent
-    tag: "1.1.1"
+    tag: "1.2.0"
 
   extraEnvs: []
 
@@ -150,7 +153,7 @@ webserver:
   image:
     repository: "allegroai/clearml"
     pullPolicy: IfNotPresent
-    tag: "1.1.1"
+    tag: "1.2.0"
 
   podAnnotations: {}
 
@@ -291,6 +294,43 @@ agentGroups:
 
     affinity: {}
 
+# This agent will spawn queued experiments in new pods, a good use case is to combine this with
+# GPU autoscaling nodes.
+# https://github.com/allegroai/clearml-agent/tree/master/docker/k8s-glue
+agentk8sglue:
+  enabled: false
+  image:
+    repository: "allegroai/clearml-agent-k8s"
+    tag: "aws-latest-1.21"
+  serviceAccountName: default
+  maxPods: 10
+  defaultDockerImage: nvidia/cuda:11.3.1-cudnn8-runtime-ubuntu20.04   # default docker image that is spawned as new pod
+  queue: aws-instances    # create this queue manually in the UI first for it to work
+  id: k8s-agent
+  podTemplate:
+    volumes: []
+      # - name: "yourvolume"
+      #   path: "/yourpath"
+    env: []
+      # # to setup access to private repo, setup secret with git credentials:
+      # - name: CLEARML_AGENT_GIT_USER
+      #   value: mygitusername
+      # - name: CLEARML_AGENT_GIT_PASS
+      #   valueFrom:
+      #     secretKeyRef:
+      #       name: git-password
+      #       key: git-password
+    resources: {}
+      # limits:
+      #   nvidia.com/gpu: 1
+    tolerations: []
+      # - key: "nvidia.com/gpu"
+      #   operator: Exists
+      #   effect: "NoSchedule"
+    nodeSelector: {}
+      # fleet: gpu-nodes
+
+
 externalServices:
   # -- Existing ElasticSearch Hostname to use if elasticsearch.enabled is false
   elasticsearchHost: ""
@@ -337,7 +377,7 @@ mongodb:  # configuration from https://github.com/bitnami/charts/blob/master/bit
     port: 27017
     portName: mongo-service
 
-elasticsearch:  # configuration from https://github.com/elastic/helm-charts/blob/7.10/elasticsearch/values.yaml
+elasticsearch:  # configuration from https://github.com/elastic/helm-charts/blob/7.16/elasticsearch/values.yaml
   enabled: true
   httpPort: 9200
   roles: