feat(service): add hpa and pdb

Dockerfile

@@ -2,7 +2,7 @@ FROM python:3.9-slim as base
 
 # hadolint ignore=DL3008,DL3009,DL3013
 RUN apt-get update && \
-    apt-get install --no-install-recommends -y git git-lfs=2.* python3-dev && \
+    apt-get install --no-install-recommends -y git git-lfs=2.* python3-dev tini && \
     pip install --no-cache-dir --upgrade pip poetry
 
 FROM base as builder
@@ -52,4 +52,5 @@ ENV RENKU_SVC_NUM_WORKERS 4
 ENV RENKU_SVC_NUM_THREADS 8
 ENV RENKU_DISABLE_VERSION_CHECK=1
 
-ENTRYPOINT ["renku"]
+ENTRYPOINT ["tini", "-g", "--"]
+CMD ["renku"]

helm-chart/renku-core/templates/deployment.yaml

@@ -17,7 +17,10 @@ metadata:
 spec:
   replicas: {{ $.Values.replicaCount }}
   strategy:
-    type: Recreate
+    type: RollingUpdate
+    rollingUpdate:
+      maxUnavailable: 0
+      maxSurge: 1
   selector:
     matchLabels:
       app.kubernetes.io/name: {{ include "renku-core.name" $ }}
@@ -165,7 +168,7 @@ spec:
               path: /health
               port: http
           resources:
-            {{- toYaml $.Values.resources | nindent 12 }}
+            {{- toYaml $.Values.resources.core | nindent 12 }}
         - name: {{ $.Chart.Name }}-datasets-workers
           image: "{{ $version.image.repository }}:{{ $version.image.tag }}"
           imagePullPolicy: {{ $version.image.pullPolicy }}
@@ -221,6 +224,8 @@ spec:
             - name: shared-volume
               mountPath: {{ $.Values.cacheDirectory }}
             {{- include "certificates.volumeMounts.system" $ | nindent 12 }}
+          resources:
+            {{- toYaml $.Values.resources.datasets-workers | nindent 12 }}
 
         - name: {{ $.Chart.Name }}-management-workers
           image: "{{ $version.image.repository }}:{{ $version.image.tag }}"
@@ -275,6 +280,8 @@ spec:
             - name: shared-volume
               mountPath: {{ $.Values.cacheDirectory }}
             {{- include "certificates.volumeMounts.system" $ | nindent 12 }}
+          resources:
+            {{- toYaml $.Values.resources.management-workers | nindent 12 }}
 
         - name: {{ $.Chart.Name }}-scheduler
           image: "{{ $version.image.repository }}:{{ $version.image.tag }}"
@@ -315,6 +322,8 @@ spec:
             {{- include "certificates.env.python" $ | nindent 12 }}
           volumeMounts:
             {{- include "certificates.volumeMounts.system" $ | nindent 12 }}
+          resources:
+            {{- toYaml $.Values.resources.scheduler | nindent 12 }}
 
         - name: traefik
           image: "{{ $.Values.traefik.image.name }}:{{ $.Values.traefik.image.tag }}"
@@ -338,6 +347,8 @@ spec:
           volumeMounts:
             - name: traefik-config
               mountPath: /config
+          resources:
+            {{- toYaml $.Values.resources.traefik | nindent 12 }}
 
       {{- with $.Values.nodeSelector }}
       nodeSelector:

helm-chart/renku-core/templates/hpa.yaml

@@ -0,0 +1,28 @@
+{{- range $version := .Values.versions }}
+{{- if .Capabilities.APIVersions.Has "autoscaling/v2" }}
+apiVersion: autoscaling/v2
+{{- else if .Capabilities.APIVersions.Has "autoscaling/v2beta2" }}
+apiVersion: autoscaling/v2beta2
+{{- else if .Capabilities.APIVersions.Has "autoscaling/v2beta1" }}
+apiVersion: autoscaling/v2beta1
+{{- else }}
+  {{- fail "ERROR: You must have at least autoscaling/v2beta1 to use HorizontalPodAutoscaler" }}
+{{- end }}
+kind: HorizontalPodAutoscaler
+metadata:
+  name: {{ include "renku-core.fullname" $ }}-{{ $version.name }}
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: {{ include "renku-core.fullname" $ }}-{{ $version.name }}
+  minReplicas: {{ $.Values.horizontalPodAutoscaling.minReplicas }}
+  maxReplicas: {{ $.Values.horizontalPodAutoscaling.maxReplicas }}
+  metrics:
+    - type: Resource
+      resource:
+        name: memory
+        target:
+          type: Utilization
+          averageUtilization: {{ $.Values.horizontalPodAutoscaling.averageMemoryUtilization }}
+{{ end }}

helm-chart/renku-core/templates/pdb.yaml

@@ -0,0 +1,11 @@
+{{- range $version := .Values.versions }}
+apiVersion: policy/v1
+kind: PodDisruptionBudget
+metadata:
+  name: {{ include "renku-core.fullname" $ }}-{{ $version.name }}
+spec:
+  minAvailable: 1
+  selector:
+    matchLabels:
+      app.kubernetes.io/deploymentVersion: {{ $version.name }}
+{{ end }}

helm-chart/renku-core/values.schema.json

@@ -141,8 +141,30 @@
           "type": "object"
       },
       "nodeSelector": {
-          "description": "Node selector for deployment",
-          "type": "object"
+        "description": "Node selector for deployment",
+        "type": "object"
+    },
+      "horizontalPodAutoscaling": {
+          "description": "Setup for scaling the core service",
+          "type": "object",
+          "properties": {
+            "minReplicas": {
+                "description": "Flag to turn on/off Sentry",
+                "type": "integer",
+                "minimum": 1
+            },
+            "maxReplicas": {
+                "description": "URI of the sentry Data Source Name",
+                "type": "integer",
+                "minimum": 1
+            },
+            "averageMemoryUtilization": {
+                "description": "Name of the sentry environment to post to",
+                "type": "integer",
+                "exclusiveMinimum": 0,
+                "exclusiveMaximum": 100
+            }
+        }
       },
       "tolerations": {
           "description": "Tolerations for deployment",

helm-chart/renku-core/values.yaml

@@ -1,7 +1,7 @@
 # Default values for renku-core.
 # This is a YAML-formatted file.
 # Declare variables to be passed into your templates.
-replicaCount: 1
+replicaCount: 2
 global:
   ## Specify a secret that containes the certificate
   ## if you would like to use a custom CA. The key for the secret
@@ -73,17 +73,25 @@ metrics:
     repository: renku/rqmetrics
     tag: 0.0.2
     pullPolicy: IfNotPresent
-resources: {}
-# We usually recommend not to specify default resources and to leave this as a conscious
-# choice for the user. This also increases chances charts run on environments with little
-# resources, such as Minikube. If you do want to specify resources, uncomment the following
-# lines, adjust them as necessary, and remove the curly braces after 'resources:'.
-# limits:
-#   cpu: 100m
-#   memory: 128Mi
-# requests:
-#   cpu: 100m
-#   memory: 128Mi
+resources:
+  core:
+    requests:
+      memory: 4Gi
+  rqmetrics:
+    requests:
+      memory: 50Mi
+  datasets-workers:
+    requests:
+      memory: 2Gi
+  management-workers:
+    requests:
+      memory: 100Mi
+  scheduler:
+    requests:
+      memory: 100Mi
+  traefik:
+    requests:
+      memory: 100Mi
 
 # nodeSelector: {}
 
@@ -129,3 +137,8 @@ traefik:
     name: renku/core-traefik
     tag: v1.8.0
     pullPolicy: IfNotPresent
+
+horizontalPodAutoscaling:
+  minReplicas: 2
+  maxReplicas: 10
+  averageMemoryUtilization: 50