Cluster variable

[IvanKolbasiuk]

Could you explain the format for
Cluster: os.Getenv(dfv1.EnvCluster)
in the controller manager code
I've got this error during setup on Azure AKS

`kubectl logs -f controller-manager-f66bd546d-w2l9f -n dev manager
time="2022-02-02T22:42:50Z" level=info msg=cpu numCPU=2
time="2022-02-02T22:42:50Z" level=info msg=process pid=1
time="2022-02-02T22:42:50Z" level=info msg=version version=v0.0.0-latest-0
time="2022-02-02T22:42:50Z" level=info msg="scaling config" defaultPeekDelay=4m0s defaultScalingDelay=1m0s
time="2022-02-02T22:42:50Z" level=info msg="reconciler config" imageFormat="quay.io/argoprojlabs/%s:latest" imagePullSecrets="[]" pullPolicy="""" runnerImage="quay.io/argoprojlabs/dataflow-runner:latest" updateInterval=15s
I0202 22:42:51.603095 1 request.go:655] Throttling request took 1.000760856s, request: GET:https://10.0.0.1:443/apis/coordination.k8s.io/v1?timeout=32s
time="2022-02-02T22:42:52Z" level=info msg="metrics server is starting to listen" addr="127.0.0.1:9090" logger=controller-runtime.metrics
panic: unable to create controller manager: cluster must be set

goroutine 1 [running]:
main.main()
/workspace/manager/main.go:97 +0xa4e`

Thanks

[alexec]

All clusters you install must be given a unique name. This is sometimes used to create UUIDs for messages.

[IvanKolbasiuk]

Thanks @alexec , I tried with setting env variable for deployment, but still receive the same error.

 - args:
    - --metrics-addr=127.0.0.1:9090
    - --enable-leader-election
    command:
    - /manager
    image: quay.io/argoprojlabs/dataflow-controller:latest
    name: manager
    ports:
    - containerPort: 9090
    resources:
      limits:
        cpu: 100m
        memory: 30Mi
      requests:
        cpu: 100m
        memory: 20Mi
    env:
    - name: Cluster
      value: "cluster name"

[IvanKolbasiuk]

Found quick_start.yaml in the release. Fixed with env: ARGO_DATAFLOW_CLUSTER: