Multi-namespace deployment

Camunda 8 Self-Managed offers flexible deployment options that allow it to span multiple namespaces. This setup consists of:

One Management Cluster: Console, Identity, and Web Modeler
One or more Orchestration Cluster: Zeebe, Operate, Tasklist, and Optimize.

For this configuration, each namespace is set up independently using Helm, with deployments categorized into two types: management and orchestration. Each type has a dedicated values file tailored to its specific deployment requirements.

Purpose

The primary goal of a multi-namespace deployment is logical grouping, which is beneficial for scenarios such as resource management. This approach treats all namespaces as part of a single logical cluster.

note

Security considerations are outside the scope of this deployment model. As a result, all credentials associated with the same deployment are shared across namespaces.

In the following sections, we demonstrate a multi-namespace Camunda deployment. One namespace is assigned to the management cluster, while two other namespaces are used for the orchestration cluster.

Management deployment

This deployment includes centralized Identity and Web Modeler only. Identity allows the other two deployments to authenticate against Keycloak. Web Modeler is a central tool for designing and deploying business process diagrams across all your automation deployments. Create the following Management deployment Helm values file.

# File: camunda-main.yaml
global:
  ingress:
    host: camunda-main.example.com
  identity:
    auth:
      connectors:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      operate:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
        redirectUrl: "https://camunda-team01.example.com/operate,https://camunda-team02.example.com/operate"
      tasklist:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
        redirectUrl: "https://camunda-team01.example.com/tasklist,https://camunda-team02.example.com/tasklist"
      optimize:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
        redirectUrl: "https://camunda-team01.example.com/optimize,https://camunda-team02.example.com/optimize"
      zeebe:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
zeebe:
  enabled: false
zeebe-gateway:
  enabled: false
operate:
  enabled: false
tasklist:
  enabled: false
optimize:
  enabled: false
connectors:
  enabled: false
elasticsearch:
  enabled: false

Install Camunda Management cluster with Helm:

helm install camunda camunda/camunda-platform --version $HELM_CHART_VERSION \
  -n camunda-main \
  -f camunda-main.yaml

Team One deployment

Let's create a Camunda automation cluster that can be owned and managed by Team One and will be deployed into namespace camunda-team01. This deployment includes Zeebe, Operate, Tasklist, and Optimize, and authenticates against Keycloak in the Management deployment:

# File: camunda-team01.yaml
global:
  ingress:
    host: camunda-team01.example.com
  identity:
    auth:
      publicIssuerUrl: "https://camunda-main.example.com/auth/realms/camunda-platform"
      connectors:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      operate:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      tasklist:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      optimize:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      zeebe:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
    service:
      url: "http://camunda-main-identity.camunda-main.svc.cluster.local:80/identity"
    keycloak:
      url:
        protocol: "http"
        host: "camunda-main-keycloak.camunda-main.svc.cluster.local"
        port: "80"
identity:
  enabled: false
webModeler:
  enabled: false
postgresql:
  enabled: false

Then, install as usual:

helm template camunda camunda/camunda-platform \
  -n camunda-team01 \
  -f camunda-team01.yaml

Team Two deployment

This deployment uses namespace camunda-team02, includes Zeebe, Operate, Tasklist, and Optimize, and authenticates against Keycloak in the management deployment:

# File: camunda-team02.yaml
global:
  ingress:
    host: camunda-team02.example.com
  identity:
    auth:
      publicIssuerUrl: "https://camunda-main.example.com/auth/realms/camunda-platform"
      connectors:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      operate:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      tasklist:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      optimize:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
      zeebe:
        existingSecret: <APP_MACHINE2MACHINE_SECRET>
    service:
      url: "http://camunda-main-identity.camunda-main.svc.cluster.local:80/identity"
    keycloak:
      url:
        protocol: "http"
        host: "camunda-main-keycloak.camunda-main.svc.cluster.local"
        port: "80"
identity:
  enabled: false
webModeler:
  enabled: false
postgresql:
  enabled: false

Then, install as usual:

helm install camunda camunda/camunda-platform --version $HELM_CHART_VERSION \
  -n camunda-team02 \
  -f camunda-team02.yaml

Multi-namespace Camunda Console deployment

To use Camunda Console Self-Managed in a multi-namespace setup, obtain Helm deployment configuration for all namespaces/deployments you would like Console to manage.

Update Management deployment to deploy Console Self-Managed. For more details, visit this Installation guide.

Assuming Camunda clusters have been deployed using the above examples, run the following script to get the release information for all deployments.

DEPLOYMENTS="camunda-main camunda-team01 camunda-team02"

for DEPLOYMENT in ${DEPLOYMENTS}; do
  helm template camunda camunda/camunda-platform \
    -n ${DEPLOYMENT} -f ${DEPLOYMENT}.yaml \
    --show-only templates/camunda/configmap-release.yaml |
      kubectl apply --dry-run=client -o jsonpath='{.data.info}' -f -
done

Add the output of this command to the management deployment values file camunda-main.yaml:

console:
  configuration: |
    camunda:
      console:
        managed:
          method: plain
          releases:
          # This is the output of the script above.
          - name: camunda
            namespace: camunda-main
            version: 9.1.2
            components:
            - name: Keycloak
              url: http://camunda-main.example.com/auth
            - name: Identity
              url: http://camunda-main.example.com
              readiness: http://camunda-identity.camunda-main:80/actuator/health
              metrics: http://camunda-identity.camunda-main:82/actuator/prometheus- name: camunda
            namespace: camunda-team01
            version: 9.1.2
            components:
            - name: Operate
              url: http://camunda-team01.example.com
              readiness: http://camunda-operate.camunda-team01:80/actuator/health/readiness
              metrics: http://camunda-operate.camunda-team01:80/actuator/prometheus
            - name: Optimize
              url: http://camunda-team01.example.com
              readiness: http://camunda-optimize.camunda-team01:80/api/readyz
              metrics: http://camunda-optimize.camunda-team01:8092/actuator/prometheus
            - name: Tasklist
              url: http://camunda-team01.example.com
              readiness: http://camunda-tasklist.camunda-team01:80/actuator/health/readiness
              metrics: http://camunda-tasklist.camunda-team01:80/actuator/prometheus
            - name: Zeebe Gateway
              url: grpc://
              readiness: http://camunda-zeebe-gateway.camunda-team01:9600/actuator/health/readiness
              metrics: http://camunda-zeebe-gateway.camunda-team01:9600/actuator/prometheus
            - name: Zeebe
              readiness: http://camunda-zeebe.camunda-team01:9600/ready
              metrics: http://camunda-zeebe.camunda-team01:9600/actuator/prometheus- name: camunda
            namespace: camunda-team02
            version: 9.1.2
            components:
            - name: Operate
              url: http://camunda-team02.example.com
              readiness: http://camunda-operate.camunda-team02:80/actuator/health/readiness
              metrics: http://camunda-operate.camunda-team02:80/actuator/prometheus
            - name: Optimize
              url: http://camunda-team02.example.com
              readiness: http://camunda-optimize.camunda-team02:80/api/readyz
              metrics: http://camunda-optimize.camunda-team02:8092/actuator/prometheus
            - name: Tasklist
              url: http://camunda-team02.example.com
              readiness: http://camunda-tasklist.camunda-team02:80/actuator/health/readiness
              metrics: http://camunda-tasklist.camunda-team02:80/actuator/prometheus
            - name: Zeebe Gateway
              url: grpc://
              readiness: http://camunda-zeebe-gateway.camunda-team02:9600/actuator/health/readiness
              metrics: http://camunda-zeebe-gateway.camunda-team02:9600/actuator/prometheus
            - name: Zeebe
              readiness: http://camunda-zeebe.camunda-team02:9600/ready
              metrics: http://camunda-zeebe.camunda-team02:9600/actuator/prometheus

note

The above values highlight the changes needed to set up the multi-namespace deployment. These parameters must be incorporated/merged with other Helm Chart values (e.g., TLS or Ingress setup).

Additionally, the machine-to-machine secrets should be the same across all namespaces for each Camunda component to authenticate against Keycloak.

For more details, check the full Helm values.

Purpose​

Management deployment​

Team One deployment​

Team Two deployment​

Multi-namespace Camunda Console deployment​

Purpose

Management deployment

Team One deployment

Team Two deployment

Multi-namespace Camunda Console deployment