Multi-cluster scaling with DistributedScaledObject (ConfigMap trigger)

This guide shows how to scale one workload across multiple member clusters using DistributedScaledObject (DSO) with a kubernetes-resource trigger reading a ConfigMap value.

Prerequisites

• Kedify Agent is installed in your KEDA cluster.
• Multi-cluster member clusters are already registered in kedify-agent-multicluster-kubeconfigs.
• DistributedScaledObject controller is enabled in Kedify Agent.
• kubectl installed and access to:
  • KEDA cluster context
  • each member cluster context

Enable DSO controller if needed:

kubectl -n keda set env deploy/kedify-agent DSO_ENABLED="true"

Architecture used in this guide

• KEDA cluster runs Kedify Agent and the DistributedScaledObject resource.
• Member clusters run identical Deployment workloads.
• Trigger source is a ConfigMap in the KEDA cluster.

When ConfigMap key metric-value is higher than target value 5, DSO increases desired replicas and distributes them across member clusters by weights.

Step 1: Deploy the same workload to member clusters

Apply this manifest to each member cluster in the same namespace (default in this example):

apiVersion: apps/v1
kind: Deployment
metadata:
  name: dso-demo-app
  namespace: default
spec:
  replicas: 0
  selector:
    matchLabels:
      app: dso-demo-app
  template:
    metadata:
      labels:
        app: dso-demo-app
    spec:
      containers:
      - name: app
        image: nginx:1.27
        ports:
        - containerPort: 80

Example apply commands:

kubectl --context member-1 apply -f dso-demo-app.yaml
kubectl --context member-2 apply -f dso-demo-app.yaml

Step 2: Create trigger ConfigMap in the KEDA cluster

Create ConfigMap in the KEDA cluster (same namespace where you will create DSO):

apiVersion: v1
kind: ConfigMap
metadata:
  name: dso-mock-metric
  namespace: default
data:
  metric-value: "0"

kubectl --context keda-cluster apply -f dso-mock-metric.yaml

Step 3: Create DistributedScaledObject

Apply this DistributedScaledObject in the KEDA cluster:

apiVersion: keda.kedify.io/v1alpha1
kind: DistributedScaledObject
metadata:
  name: dso-demo
  namespace: default
spec:
  memberClusters:
  - name: member-1
    weight: 4
  - name: member-2
    weight: 6
  scaledObjectSpec:
    scaleTargetRef:
      kind: Deployment
      name: dso-demo-app
    minReplicaCount: 0
    maxReplicaCount: 10
    triggers:
    - type: kubernetes-resource
      metadata:
        resourceKind: ConfigMap
        resourceName: dso-mock-metric
        key: metric-value
        targetValue: "5"

kubectl --context keda-cluster apply -f distributedscaledobject.yaml

Step 4: Trigger scaling

Increase the metric value above the target:

kubectl --context keda-cluster -n default patch configmap dso-mock-metric \
  --type merge \
  -p '{"data":{"metric-value":"20"}}'

Set it back below target to scale down:

kubectl --context keda-cluster -n default patch configmap dso-mock-metric \
  --type merge \
  -p '{"data":{"metric-value":"0"}}'

Step 5: Verify distribution and status

Check DSO status in KEDA cluster:

kubectl --context keda-cluster -n default get distributedscaledobject dso-demo -o yaml

Check replicas in each member cluster:

kubectl --context member-1 -n default get deploy dso-demo-app
kubectl --context member-2 -n default get deploy dso-demo-app

You should observe replica distribution following weights (member-1:4, member-2:6) within rounding limits.

Troubleshooting

• DSO does not scale:
  • Verify DSO_ENABLED="true" on kedify-agent.
  • Verify member cluster names in spec.memberClusters exactly match registered member names.
  • Verify Deployment name and namespace are identical across member clusters.
• Trigger not firing:
  • Verify ConfigMap exists in the same namespace as DSO.
  • Verify metric-value is numeric and above targetValue.
• Member cluster stays at 0 replicas:
  • Check DSO status field memberClusterStatuses for connectivity or auth issues.