Deploying Data Index and multiple SonataFlow applications on Minikube

This document describes how to deploy a multiple SonataFlow workflow applications and the Data Index service using a local Kubernetes cluster, such as Minikube, using the SonataFlow Operator.

For more information about Minikube and related system requirements, see Getting started with Minikube documentation.

This use case is intended to represent an installation with:

A singleton Data Index Service with PostgreSQL persistence
The greeting workflow (no persistence), that is configured to register events to the Data Index Service.
The helloworld workflow (no persistence), that is configured to register events to the Data Index Service.
Both workflows are configured to register the process events on the Data Index Service.

You can directly access the UseCase2 example application we are going to follow at SonataFlow Data Index Use Cases with operator.

Prerequisites

Minikube installed with registry addon enabled
kubectl command-line tool is installed. Otherwise, Minikube handles it.
SonataFlow operator installed if workflows are deployed. To install the operator you can see Install the SonataFlow Operator.

We recommend that you start Minikube with the following parameters, note that the registry addon must be enabled.

minikube start --cpus 4 --memory 10240 --addons registry --addons metrics-server --insecure-registry "10.0.0.0/24" --insecure-registry "localhost:5000"

To verify that the registry addon was property added you can execute this command:

minikube addons list | grep registry

| registry                    | minikube | enabled ✅   | Google                         |
| registry-aliases            | minikube | disabled     | 3rd party (unknown)            |
| registry-creds              | minikube | disabled     | 3rd party (UPMC Enterprises)   |

You can check the Minikube installation by entering the following commands in a command terminal:

Verify Minikube version

minikube version

Verify kubectl CLI version

kubectl version

If kubectl is not installed, then Minikube handles it when you execute the following command:

kubectl is available using Minikube

alias kubectl="minikube kubectl --"

Procedure

After cloning the SonataFlow examples repository. Open a terminal and run the following commands
cd serverless-operator-examples/serverless-workflow-dataindex-use-cases/
Create the namespace:
kubectl create namespace usecase2

Deploy the Data Index Service and postgresql database:

Here you can find the infrastructure kustomization required to deploy Data Index service and a postgresql database explained in this use case.

Thas folder contains four files:

kustomization.yaml
01-postgres.yaml
02-dataindex.yaml
application.properties

kustomization.yaml resources that deploy Data Index deployment with persistence to a postgresql database

resources:
- 01-postgres.yaml (1)
- 02-dataindex.yaml (2)

secretGenerator:
  - name: postgres-secrets
    literals:
      - POSTGRES_USER=sonataflow
      - POSTGRES_PASSWORD=sonataflow
      - POSTGRES_DB=sonataflow
      - PGDATA=/var/lib/postgresql/data/mydata

configMapGenerator:
  - name: dataindex-properties
    files:
      - application.properties

1	Postgres database deployment
2	Data Index deployment

01_postgres.yaml that deploys Postgresql database

---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  labels:
    app.kubernetes.io/name: postgres
  name: postgres-pvc
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi
---
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app.kubernetes.io/name: postgres
  name: postgres
spec:
  replicas: 1
  selector:
    matchLabels:
      app.kubernetes.io/name: postgres
  template:
    metadata:
      labels:
        app.kubernetes.io/name: postgres
    spec:
      containers:
        - name: postgres
          image: postgres:13.2-alpine
          imagePullPolicy: 'IfNotPresent'
          ports:
            - containerPort: 5432
          volumeMounts:
            - name: storage
              mountPath: /var/lib/postgresql/data
          envFrom:
            - secretRef:
                name: postgres-secrets
          readinessProbe:
            exec:
              command: ["pg_isready"]
            initialDelaySeconds: 15
            timeoutSeconds: 2
          livenessProbe:
            exec:
              command: ["pg_isready"]
            initialDelaySeconds: 15
            timeoutSeconds: 2
          resources:
            limits:
              memory: "256Mi"
              cpu: "500m"
      volumes:
      - name: storage
        persistentVolumeClaim:
          claimName: postgres-pvc
---
apiVersion: v1
kind: Service
metadata:
  labels:
    app.kubernetes.io/name: postgres
  name: postgres
spec:
  selector:
    app.kubernetes.io/name: postgres
  ports:
    - port: 5432

02-dataindex.yaml that deploys Data Index with persistence to the previous defined postgresql database

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app.kubernetes.io/name: data-index-service-postgresql
  name: data-index-service-postgresql
spec:
  replicas: 1
  selector:
    matchLabels:
      app.kubernetes.io/name: data-index-service-postgresql
  template:
    metadata:
      labels:
        app.kubernetes.io/name: data-index-service-postgresql
    spec:
      containers:
        - name: data-index-service-postgresql
          image: quay.io/kiegroup/kogito-data-index-postgresql:latest
          imagePullPolicy: Always
          resources:
            limits:
              memory: "256Mi"
              cpu: "500m"
          ports:
            - containerPort: 8080
              name: http
              protocol: TCP
          env:
            - name: KOGITO_DATA_INDEX_QUARKUS_PROFILE
              value: http-events-support
            - name: KUBERNETES_NAMESPACE
              valueFrom:
                fieldRef:
                  fieldPath: metadata.namespace
            - name: QUARKUS_DATASOURCE_USERNAME
              valueFrom:
                secretKeyRef:
                  key: POSTGRES_USER
                  name: postgres-secrets
            - name: QUARKUS_DATASOURCE_PASSWORD
              valueFrom:
                secretKeyRef:
                  key: POSTGRES_PASSWORD
                  name: postgres-secrets
          volumeMounts:
            - name: application-config
              mountPath: "/home/kogito/config"
          livenessProbe:
            failureThreshold: 3
            httpGet:
              path: /q/health/live
              port: 8080
              scheme: HTTP
            initialDelaySeconds: 0
            periodSeconds: 30
            successThreshold: 1
            timeoutSeconds: 10
          readinessProbe:
            failureThreshold: 3
            httpGet:
              path: /q/health/ready
              port: 8080
              scheme: HTTP
            initialDelaySeconds: 0
            periodSeconds: 30
            successThreshold: 1
            timeoutSeconds: 10
      volumes:
        - name: application-config
          configMap:
            name: dataindex-properties
      initContainers:
        - name: init-postgres
          image: registry.access.redhat.com/ubi9/ubi-minimal:latest
          imagePullPolicy: IfNotPresent
          command: ['sh', '-c', 'until (echo 1 > /dev/tcp/postgres.$(cat /var/run/secrets/kubernetes.io/serviceaccount/namespace).svc.cluster.local/5432) >/dev/null 2>&1; do echo "Waiting for postgres server"; sleep 3; done;']
---
apiVersion: v1
kind: Service
metadata:
  labels:
    app.kubernetes.io/name: data-index-service-postgresql
  name: data-index-service-postgresql
spec:
  ports:
    - name: http
      port: 80
      targetPort: 8080
  selector:
    app.kubernetes.io/name: data-index-service-postgresql
  type: NodePort

application.properties referenced by kustomization.yaml

quarkus.http.port=8080
quarkus.http.cors=true
quarkus.http.cors.origins=/.*/

quarkus.datasource.jdbc.url=jdbc:postgresql://postgres:5432/sonataflow?currentSchema=data-index-service
quarkus.hibernate-orm.database.generation=update
quarkus.flyway.migrate-at-start=true

# Disable kafka client health check since the quarkus-http connector is being used instead.
quarkus.smallrye-health.check."io.quarkus.kafka.client.health.KafkaHealthCheck".enabled=false

Perform the deployments executing

kubectl kustomize infra/dataindex | kubectl apply -f - -n usecase2

configmap/dataindex-properties-hg9ff8bff5 created
secret/postgres-secrets-22tkgc2dt7 created
service/data-index-service-postgresql created
service/postgres created
persistentvolumeclaim/postgres-pvc created
deployment.apps/data-index-service-postgresql created
deployment.apps/postgres created

Give some time for the data index to start, you can check that it’s running by executing.

kubectl get pod -n usecase2

NAME                                             READY   STATUS    RESTARTS       AGE
data-index-service-postgresql-5d76dc4468-lb259   1/1     Running   0              2m11s
postgres-7f78499688-lc8n6                        1/1     Running   0              2m11s

Deploy the workflow:

Here you can find the use case kustomization required to deploy the workflow

Use case kustomization.yaml resources that deploys the workflow

resources:
- ../../infra/service_discovery
- ../../workflows/sonataflow-greeting
- ../../workflows/sonataflow-helloworld

To see in more detail access to Building and Deploying Workflows with the Operator

Perform the deployment executing

 kubectl kustomize usecases/usecase2 | kubectl apply -f - -n usecase2

configmap/greeting-props created
configmap/helloworld-props created
sonataflow.sonataflow.org/greeting created
sonataflow.sonataflow.org/helloworld created

Give some time for the sonataflow operator to build and deploy the workflow. To check that the workflow is ready you can use this command.

kubectl get workflow -n usecase2

NAME       PROFILE   VERSION   URL   READY   REASON
greeting             0.0.1           True
helloworld           0.0.1           True

Expose the workflows and get the urls:

kubectl patch svc greeting helloworld -p '{"spec": {"type": "NodePort"}}' -n usecase2

minikube service greeting --url -n usecase2

minikube service helloworld --url -n usecase2

Create a workflow instance:

You must use the URLs calculated in step 5.

curl -X POST -H 'Content-Type:application/json' -H 'Accept:application/json' -d '{"name": "John", "language": "English"}'    http://192.168.49.2:32407/greeting

curl -X POST -H 'Content-Type:application/json' -H 'Accept:application/json' -d '{}'    http://192.168.49.2:32327/helloworld

Clean the use case:
kubectl delete namespace usecase2

Querying Data Index service on Minikube

You can use the public Data Index endpoint to play around with the GraphiQL interface.

Procedure

This procedure apply to all use cases with that deploys the Data Index Service.

Get the Data Index Url:

minikube service data-index-service-postgresql --url -n my_usecase

Open the GrahiqlUI

Using the url returned, open a browser window in the following url http://192.168.49.2:32409/graphiql/,

that IP and port will be different in your installation, and don’t forget to add the last slash "/" to the url, otherwise the GraphiqlUI won’t be opened.

To see the process instances information you can execute this query:

{
  ProcessInstances {
    id,
    processId,
    processName,
    variables,
    state,
    endpoint,
    serviceUrl,
    start,
    end
  }
}

The results should be something like:

{
  "data": {
    "ProcessInstances": [
      {
        "id": "3ed8bf63-85c9-425d-9099-49bfb63608cb",
        "processId": "greeting",
        "processName": "workflow",
        "variables": "{\"workflowdata\":{\"name\":\"John\",\"greeting\":\"Hello from JSON Workflow, \",\"language\":\"English\"}}",
        "state": "COMPLETED",
        "endpoint": "/greeting",
        "serviceUrl": "http://greeting",
        "start": "2023-09-13T06:59:24.319Z",
        "end": "2023-09-13T06:59:24.400Z"
      }
    ]
  }
}

To see the jobs instances information, if any, you can execute this query:

{
  Jobs {
    id,
    processId,
    processInstanceId,
    status,
    expirationTime,
    retries,
    endpoint,
    callbackEndpoint
  }
}

The results should be something like:

{
  "data": {
    "Jobs": [
      {
        "id": "55c7aadb-3dff-4b97-af8e-cc45014b1c0d",
        "processId": "callbackstatetimeouts",
        "processInstanceId": "299886b7-2b78-4965-a701-16783c4162d8",
        "status": "EXECUTED",
        "expirationTime": null,
        "retries": 0,
        "endpoint": "http://jobs-service-postgresql/jobs",
        "callbackEndpoint": "http://callbackstatetimeouts:80/management/jobs/callbackstatetimeouts/instances/299886b7-2b78-4965-a701-16783c4162d8/timers/-1"
      }
    ]
  }
}

Additional resources

Found an issue?

If you find an issue or any misleading information, please feel free to report it here. We really appreciate it!