Merge pull request #12564 from Budibase/ops-228-split-out-automations-from-main-app-infrastructure

Split out automation processing from the apps pods.

charts/budibase/README.md

@@ -157,6 +157,17 @@ $ helm install --create-namespace --namespace budibase budibase . -f values.yaml
 | services.apps.replicaCount | int | `1` | The number of apps replicas to run. |
 | services.apps.resources | object | `{}` | The resources to use for apps pods. See <https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/> for more information on how to set these. |
 | services.apps.startupProbe | object | HTTP health checks. | Startup probe configuration for apps pods. You shouldn't need to change this, but if you want to you can find more information here: <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/> |
+| services.automationWorkers.autoscaling.enabled | bool | `false` | Whether to enable horizontal pod autoscaling for the apps service. |
+| services.automationWorkers.autoscaling.maxReplicas | int | `10` |  |
+| services.automationWorkers.autoscaling.minReplicas | int | `1` |  |
+| services.automationWorkers.autoscaling.targetCPUUtilizationPercentage | int | `80` | Target CPU utilization percentage for the automation worker service. Note that for autoscaling to work, you will need to have metrics-server configured, and resources set for the automation worker pods. |
+| services.automationWorkers.enabled | bool | `true` | Whether or not to enable the automation worker service. If you disable this, automations will be processed by the apps service. |
+| services.automationWorkers.livenessProbe | object | HTTP health checks. | Liveness probe configuration for automation worker pods. You shouldn't need to change this, but if you want to you can find more information here: <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/> |
+| services.automationWorkers.logLevel | string | `"info"` | The log level for the automation worker service. |
+| services.automationWorkers.readinessProbe | object | HTTP health checks. | Readiness probe configuration for automation worker pods. You shouldn't need to change this, but if you want to you can find more information here: <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/> |
+| services.automationWorkers.replicaCount | int | `1` | The number of automation worker replicas to run. |
+| services.automationWorkers.resources | object | `{}` | The resources to use for automation worker pods. See <https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/> for more information on how to set these. |
+| services.automationWorkers.startupProbe | object | HTTP health checks. | Startup probe configuration for automation worker pods. You shouldn't need to change this, but if you want to you can find more information here: <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/> |
 | services.couchdb.backup.enabled | bool | `false` | Whether or not to enable periodic CouchDB backups. This works by replicating to another CouchDB instance. |
 | services.couchdb.backup.interval | string | `""` | Backup interval in seconds |
 | services.couchdb.backup.resources | object | `{}` | The resources to use for CouchDB backup pods. See <https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/> for more information on how to set these. |

charts/budibase/templates/app-service-deployment.yaml

@@ -192,7 +192,10 @@ spec:
         - name: NODE_TLS_REJECT_UNAUTHORIZED
           value: {{ .Values.services.tlsRejectUnauthorized }}
         {{ end }}
-
+        {{- if .Values.services.automationWorkers.enabled }}
+        - name: APP_FEATURES
+          value: "api"
+        {{- end }}
         image: budibase/apps:{{ .Values.globals.appVersion | default .Chart.AppVersion }}
         imagePullPolicy: Always
         {{- if .Values.services.apps.startupProbe }}

charts/budibase/templates/automation-worker-service-deployment.yaml

@@ -0,0 +1,244 @@
+{{- if .Values.services.automationWorkers.enabled }}
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  annotations:
+{{ if .Values.services.automationWorkers.deploymentAnnotations }}
+{{- toYaml .Values.services.automationWorkers.deploymentAnnotations | indent 4 -}}
+{{ end }}
+  labels:
+    io.kompose.service: automation-worker-service
+{{ if .Values.services.automationWorkers.deploymentLabels }}
+{{- toYaml .Values.services.automationWorkers.deploymentLabels | indent 4 -}}
+{{ end }}
+  name: automation-worker-service
+spec:
+  replicas: {{ .Values.services.automationWorkers.replicaCount }}
+  selector:
+    matchLabels:
+      io.kompose.service: automation-worker-service
+  strategy:
+    type: RollingUpdate
+  template:
+    metadata:
+      annotations:
+{{ if .Values.services.automationWorkers.templateAnnotations }}
+{{- toYaml .Values.services.automationWorkers.templateAnnotations | indent 8 -}}
+{{ end }}
+      labels:
+        io.kompose.service: automation-worker-service
+{{ if .Values.services.automationWorkers.templateLabels }}
+{{- toYaml .Values.services.automationWorkers.templateLabels | indent 8 -}}
+{{ end }}
+    spec:
+      containers:
+      - env:
+        - name: BUDIBASE_ENVIRONMENT
+          value: {{ .Values.globals.budibaseEnv }}
+        - name: DEPLOYMENT_ENVIRONMENT
+          value: "kubernetes"
+        - name: COUCH_DB_URL
+          {{ if .Values.services.couchdb.url }}
+          value: {{ .Values.services.couchdb.url }}
+          {{ else }}
+          value: http://{{ .Release.Name }}-svc-couchdb:{{ .Values.services.couchdb.port }}
+          {{ end }}
+        {{ if .Values.services.couchdb.enabled }}
+        - name: COUCH_DB_USER
+          valueFrom:
+            secretKeyRef:
+              name: {{ template "couchdb.fullname" . }}
+              key: adminUsername
+        - name: COUCH_DB_PASSWORD
+          valueFrom:
+            secretKeyRef:
+              name: {{ template "couchdb.fullname" . }}
+              key: adminPassword
+        {{ end }}
+        - name: ENABLE_ANALYTICS
+          value: {{ .Values.globals.enableAnalytics | quote }}
+        - name: API_ENCRYPTION_KEY
+          value: {{ .Values.globals.apiEncryptionKey | quote }}
+        - name: HTTP_LOGGING
+          value: {{ .Values.services.automationWorkers.httpLogging | quote }}
+        - name: INTERNAL_API_KEY
+          valueFrom:
+            secretKeyRef:
+              name: {{ template "budibase.fullname" . }}
+              key: internalApiKey
+        - name: INTERNAL_API_KEY_FALLBACK
+          value: {{ .Values.globals.internalApiKeyFallback | quote }}
+        - name: JWT_SECRET
+          valueFrom:
+            secretKeyRef:
+              name: {{ template "budibase.fullname" . }}
+              key: jwtSecret
+        - name: JWT_SECRET_FALLBACK
+          value: {{ .Values.globals.jwtSecretFallback | quote }}
+        {{ if .Values.services.objectStore.region }}
+        - name: AWS_REGION
+          value: {{ .Values.services.objectStore.region }}
+        {{ end }}
+        - name: MINIO_ENABLED
+          value: {{ .Values.services.objectStore.minio | quote }}
+        - name: MINIO_ACCESS_KEY
+          valueFrom:
+            secretKeyRef:
+              name: {{ template "budibase.fullname" . }}
+              key: objectStoreAccess
+        - name: MINIO_SECRET_KEY
+          valueFrom:
+            secretKeyRef:
+              name: {{ template "budibase.fullname" . }}
+              key: objectStoreSecret
+        - name: CLOUDFRONT_CDN
+          value: {{ .Values.services.objectStore.cloudfront.cdn | quote }}
+        - name: CLOUDFRONT_PUBLIC_KEY_ID
+          value: {{ .Values.services.objectStore.cloudfront.publicKeyId | quote }}
+        - name: CLOUDFRONT_PRIVATE_KEY_64
+          value: {{ .Values.services.objectStore.cloudfront.privateKey64 | quote }}
+        - name: MINIO_URL
+          value: {{ .Values.services.objectStore.url }}
+        - name: PLUGIN_BUCKET_NAME
+          value: {{ .Values.services.objectStore.pluginBucketName | quote }}
+        - name: APPS_BUCKET_NAME
+          value: {{ .Values.services.objectStore.appsBucketName | quote }}
+        - name: GLOBAL_BUCKET_NAME
+          value: {{ .Values.services.objectStore.globalBucketName | quote }}
+        - name: BACKUPS_BUCKET_NAME
+          value: {{ .Values.services.objectStore.backupsBucketName | quote }}
+        - name: PORT
+          value: {{ .Values.services.automationWorkers.port | quote }}
+        {{ if .Values.services.worker.publicApiRateLimitPerSecond }}
+        - name: API_REQ_LIMIT_PER_SEC
+          value: {{ .Values.globals.automationWorkers.publicApiRateLimitPerSecond | quote }}
+        {{ end }}
+        - name: MULTI_TENANCY
+          value: {{ .Values.globals.multiTenancy | quote }}
+        - name: OFFLINE_MODE
+          value: {{ .Values.globals.offlineMode | quote }}
+        - name: LOG_LEVEL
+          value: {{ .Values.services.automationWorkers.logLevel | quote }}
+        - name: REDIS_PASSWORD
+          value: {{ .Values.services.redis.password }}
+        - name: REDIS_URL
+          {{ if .Values.services.redis.url }}
+          value: {{ .Values.services.redis.url }}
+          {{ else }}
+          value: redis-service:{{ .Values.services.redis.port }}
+          {{ end }}
+        - name: SELF_HOSTED
+          value: {{ .Values.globals.selfHosted | quote }}
+        - name: POSTHOG_TOKEN
+          value: {{ .Values.globals.posthogToken | quote }}
+        - name: WORKER_URL
+          value: http://worker-service:{{ .Values.services.worker.port }}
+        - name: PLATFORM_URL
+          value: {{ .Values.globals.platformUrl | quote }}
+        - name: ACCOUNT_PORTAL_URL
+          value: {{ .Values.globals.accountPortalUrl | quote }}
+        - name: ACCOUNT_PORTAL_API_KEY
+          value: {{ .Values.globals.accountPortalApiKey | quote }}
+        - name: COOKIE_DOMAIN
+          value: {{ .Values.globals.cookieDomain | quote }}
+        - name: HTTP_MIGRATIONS
+          value: {{ .Values.globals.httpMigrations | quote }}
+        - name: GOOGLE_CLIENT_ID
+          value: {{ .Values.globals.google.clientId | quote }}
+        - name: GOOGLE_CLIENT_SECRET
+          value: {{ .Values.globals.google.secret | quote }}
+        - name: AUTOMATION_MAX_ITERATIONS
+          value: {{ .Values.globals.automationMaxIterations | quote }}
+        - name: TENANT_FEATURE_FLAGS
+          value: {{ .Values.globals.tenantFeatureFlags | quote }}
+        - name: ENCRYPTION_KEY
+          value: {{ .Values.globals.bbEncryptionKey | quote }}
+        {{ if .Values.globals.bbAdminUserEmail }}
+        - name: BB_ADMIN_USER_EMAIL
+          value: {{ .Values.globals.bbAdminUserEmail | quote }}
+        {{ end }}
+        {{ if .Values.globals.bbAdminUserPassword }}
+        - name: BB_ADMIN_USER_PASSWORD
+          value: {{ .Values.globals.bbAdminUserPassword | quote }}
+        {{ end }}
+        {{ if .Values.globals.pluginsDir }}
+        - name: PLUGINS_DIR
+          value: {{ .Values.globals.pluginsDir | quote }}
+        {{ end }}
+        {{ if .Values.services.automationWorkers.nodeDebug }}
+        - name: NODE_DEBUG
+          value: {{ .Values.services.automationWorkers.nodeDebug | quote }}
+        {{ end }}
+        {{ if .Values.globals.datadogApmEnabled }}
+        - name: DD_LOGS_INJECTION
+          value: {{ .Values.globals.datadogApmEnabled | quote }}
+        - name: DD_APM_ENABLED
+          value: {{ .Values.globals.datadogApmEnabled | quote }}
+        - name: DD_APM_DD_URL
+          value: https://trace.agent.datadoghq.eu
+        {{ end }}
+        {{ if .Values.globals.globalAgentHttpProxy }}
+        - name: GLOBAL_AGENT_HTTP_PROXY
+          value: {{ .Values.globals.globalAgentHttpProxy | quote }}
+        {{ end }}
+        {{ if .Values.globals.globalAgentHttpsProxy }}
+        - name: GLOBAL_AGENT_HTTPS_PROXY
+          value: {{ .Values.globals.globalAgentHttpsProxy | quote }}
+        {{ end }}
+        {{ if .Values.globals.globalAgentNoProxy }}
+        - name: GLOBAL_AGENT_NO_PROXY
+          value: {{ .Values.globals.globalAgentNoProxy | quote }}
+        {{ end }}
+        {{ if .Values.services.tlsRejectUnauthorized }}
+        - name: NODE_TLS_REJECT_UNAUTHORIZED
+          value: {{ .Values.services.tlsRejectUnauthorized }}
+        {{ end }}
+        - name: APP_FEATURES
+          value: "automations"
+
+        image: budibase/apps:{{ .Values.globals.appVersion | default .Chart.AppVersion }}
+        imagePullPolicy: Always
+        {{- if .Values.services.automationWorkers.startupProbe }}
+        {{- with .Values.services.automationWorkers.startupProbe }}
+        startupProbe:
+            {{- toYaml . | nindent 10 }}
+        {{- end }}
+        {{- end }}
+        {{- if .Values.services.automationWorkers.livenessProbe }}
+        {{- with .Values.services.automationWorkers.livenessProbe }}
+        livenessProbe:
+            {{- toYaml . | nindent 10 }}
+        {{- end }}
+        {{- end }}
+        {{- if .Values.services.automationWorkers.readinessProbe }}
+        {{- with .Values.services.automationWorkers.readinessProbe }}
+        readinessProbe:
+            {{- toYaml . | nindent 10 }}
+        {{- end }}
+        {{- end }}
+        name: bbapps
+        ports:
+        - containerPort: {{ .Values.services.automationWorkers.port }}
+        {{ with .Values.services.automationWorkers.resources }}
+        resources:
+        {{- toYaml . | nindent 10 }}
+        {{ end }}
+    {{- with .Values.affinity }}
+      affinity:
+        {{- toYaml . | nindent 8 }}
+    {{- end }}
+    {{- with .Values.tolerations }}
+      tolerations:
+        {{- toYaml . | nindent 8 }}
+    {{- end }}
+      {{ if .Values.schedulerName }}
+      schedulerName: {{ .Values.schedulerName | quote }}
+      {{ end }}    
+      {{ if .Values.imagePullSecrets }}
+      imagePullSecrets:
+      {{- toYaml .Values.imagePullSecrets | nindent 6 }}
+      {{ end }}
+      restartPolicy: Always
+      serviceAccountName: ""
+status: {}
+{{- end }}

charts/budibase/templates/automation-worker-service-hpa.yaml

@@ -0,0 +1,32 @@
+{{- if .Values.services.automationWorkers.autoscaling.enabled }}
+apiVersion: {{ ternary "autoscaling/v2" "autoscaling/v2beta2" (.Capabilities.APIVersions.Has "autoscaling/v2") }}
+kind: HorizontalPodAutoscaler
+metadata:
+  name: {{ include "budibase.fullname" . }}-apps
+  labels:
+    {{- include "budibase.labels" . | nindent 4 }}
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: automation-worker-service
+  minReplicas: {{ .Values.services.automationWorkers.autoscaling.minReplicas }}
+  maxReplicas: {{ .Values.services.automationWorkers.autoscaling.maxReplicas }}
+  metrics:
+    {{- if .Values.services.automationWorkers.autoscaling.targetCPUUtilizationPercentage }}
+    - type: Resource
+      resource:
+        name: cpu
+        target:
+          type: Utilization
+          averageUtilization: {{ .Values.services.automationWorkers.autoscaling.targetCPUUtilizationPercentage }}
+    {{- end }}
+    {{- if .Values.services.automationWorkers.autoscaling.targetMemoryUtilizationPercentage }}
+    - type: Resource
+      resource:
+        name: memory
+        target:
+          type: Utilization
+          averageUtilization: {{ .Values.services.automationWorkers.autoscaling.targetMemoryUtilizationPercentage }}
+    {{- end }}
+{{- end }}

charts/budibase/values.yaml

@@ -272,6 +272,75 @@ services:
       # and resources set for the apps pods.
       targetCPUUtilizationPercentage: 80
 
+  automationWorkers:
+    # -- Whether or not to enable the automation worker service. If you disable this,
+    # automations will be processed by the apps service.
+    enabled: true
+    # @ignore (you shouldn't need to change this)
+    port: 4002
+    # -- The number of automation worker replicas to run.
+    replicaCount: 1
+    # -- The log level for the automation worker service.
+    logLevel: info
+    # -- The resources to use for automation worker pods. See
+    # <https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/>
+    # for more information on how to set these.
+    resources: {}
+    # -- Startup probe configuration for automation worker pods. You shouldn't
+    # need to change this, but if you want to you can find more information
+    # here:
+    # <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/>
+    # @default -- HTTP health checks.
+    startupProbe:
+      # @ignore
+      httpGet:
+        path: /health
+        port: 4002
+        scheme: HTTP
+      # @ignore
+      failureThreshold: 30
+      # @ignore
+      periodSeconds: 3
+    # -- Readiness probe configuration for automation worker pods. You shouldn't
+    # need to change this, but if you want to you can find more information
+    # here:
+    # <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/>
+    # @default -- HTTP health checks.
+    readinessProbe:
+      # @ignore
+      httpGet:
+        path: /health
+        port: 4002
+        scheme: HTTP
+      # @ignore
+      periodSeconds: 3
+      # @ignore
+      failureThreshold: 1
+    # -- Liveness probe configuration for automation worker pods. You shouldn't
+    # need to change this, but if you want to you can find more information
+    # here:
+    # <https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/>
+    # @default -- HTTP health checks.
+    livenessProbe:
+      # @ignore
+      httpGet:
+        path: /health
+        port: 4002
+        scheme: HTTP
+      # @ignore
+      failureThreshold: 3
+      # @ignore
+      periodSeconds: 30
+    autoscaling:
+      # -- Whether to enable horizontal pod autoscaling for the apps service.
+      enabled: false
+      minReplicas: 1
+      maxReplicas: 10
+      # -- Target CPU utilization percentage for the automation worker service.
+      # Note that for autoscaling to work, you will need to have metrics-server
+      # configured, and resources set for the automation worker pods.
+      targetCPUUtilizationPercentage: 80
+
   worker:
     # @ignore (you shouldn't need to change this)
     port: 4003

packages/backend-core/src/queue/inMemoryQueue.ts

@@ -68,6 +68,10 @@ class InMemoryQueue {
     })
   }
 
+  async isReady() {
+    return true
+  }
+
   // simply puts a message to the queue and emits to the queue for processing
   /**
    * Simple function to replicate the add message functionality of Bull, putting

packages/server/src/api/index.ts

@@ -4,15 +4,24 @@ import currentApp from "../middleware/currentapp"
 import zlib from "zlib"
 import { mainRoutes, staticRoutes, publicRoutes } from "./routes"
 import { middleware as pro } from "@budibase/pro"
-import { apiEnabled } from "../features"
+import { apiEnabled, automationsEnabled } from "../features"
 import migrations from "../middleware/appMigrations"
+import { automationQueue } from "../automations"
 
 export { shutdown } from "./routes/public"
 const compress = require("koa-compress")
 
 export const router: Router = new Router()
 
-router.get("/health", ctx => (ctx.status = 200))
+router.get("/health", async ctx => {
+  if (automationsEnabled()) {
+    if (!(await automationQueue.isReady())) {
+      ctx.status = 503
+      return
+    }
+  }
+  ctx.status = 200
+})
 router.get("/version", ctx => (ctx.body = envCore.VERSION))
 
 router.use(middleware.errorHandling)