Install a NebulaGraph cluster using NebulaGraph Operator¶

Using NebulaGraph Operator to install NebulaGraph clusters enables automated cluster management with automatic error recovery. This topic covers two methods, kubectl apply and helm, for installing clusters using NebulaGraph Operator.

Historical version compatibility

NebulaGraph Operator versions 1.x are not compatible with NebulaGraph versions below 3.x.

Prerequisites¶

Use `kubectl apply`¶

Create a namespace for storing NebulaGraph cluster-related resources. For example, create the nebula namespace.
```
kubectl create namespace nebula
```

Create a YAML configuration file nebulacluster.yaml for the cluster. For example, create a cluster named nebula.

Expand to view an example configuration for the nebula cluster

apiVersion: apps.nebula-graph.io/v1alpha1
kind: NebulaCluster
metadata:
  name: nebula
  namespace: default
spec:
  # Control the Pod scheduling strategy.
  topologySpreadConstraints:
  - topologyKey: "kubernetes.io/hostname"
    whenUnsatisfiable: "ScheduleAnyway"
  # Enable PV recycling.
  enablePVReclaim: false
  # Enable monitoring.
  exporter:
    image: vesoft/nebula-stats-exporter
    version: v3.3.0
    replicas: 1
    maxRequests: 20
  # Custom Agent image for cluster backup and restore, and log cleanup.
  agent:
    image: vesoft/nebula-agent
    version: latest
    resources:
      requests:
        cpu: "100m"
        memory: "128Mi"
      limits:
        cpu: "200m"
        memory: "256Mi"  
  # Configure the image pull policy.
  imagePullPolicy: Always
  # Select the nodes for Pod scheduling.
  nodeSelector:
    nebula: cloud
  # Dependent controller name.
  reference:
    name: statefulsets.apps
    version: v1
  # Scheduler name.
  schedulerName: default-scheduler   
  # Start NebulaGraph Console service for connecting to the Graph service.
  console:
    image: vesoft/nebula-console
    version: nightly
    username: "demo"
    password: "test"                 
  # Graph service configuration. 
  graphd:
  # Used to check if the Graph service is running normally.
  #  readinessProbe:
  #    failureThreshold: 3
  #    httpGet:
  #      path: /status
  #      port: 19669
  #      scheme: HTTP
  #    initialDelaySeconds: 40
  #    periodSeconds: 10
  #    successThreshold: 1
  #    timeoutSeconds: 10
    # Container image for the Graph service.
    image: vesoft/nebula-graphd
    logVolumeClaim:
      resources:
        requests:
          storage: 2Gi
      # Storage class name for storing Graph service logs.
      storageClassName: local-sc
    # Number of replicas for the Graph service Pod.
    replicas: 1
    # Resource configuration for the Graph service.
    resources:
      limits:
        cpu: "1"
        memory: 1Gi
      requests:
        cpu: 500m
        memory: 500Mi
    # Version of the Graph service.
    version: v3.6.0
    # Custom flags configuration for the Graph service.
    config: {}
  # Meta service configuration.
  metad:
  #  readinessProbe:
  #    failureThreshold: 3
  #    httpGet:
  #      path: /status
  #      port: 19559
  #      scheme: HTTP
  #    initialDelaySeconds: 5
  #    periodSeconds: 5
  #    successThreshold: 1
  #    timeoutSeconds: 5
    # Container image for the Meta service.
    image: vesoft/nebula-metad
    logVolumeClaim:
      resources:
        requests:
          storage: 2Gi
      storageClassName: local-sc
    dataVolumeClaim:
      resources:
        requests:
          storage: 2Gi
      storageClassName: local-sc
    replicas: 1
    resources:
      limits:
        cpu: "1"
        memory: 1Gi
      requests:
        cpu: 500m
        memory: 500Mi
    version: v3.6.0
    # Custom flags configuration for the Meta service.
    config: {}          
  # Storage service configuration.
  storaged:
  #  readinessProbe:
  #    failureThreshold: 3
  #    httpGet:
  #      path: /status
  #      port: 19779
  #      scheme: HTTP
  #    initialDelaySeconds: 40
  #    periodSeconds: 10
  #    successThreshold: 1
  #    timeoutSeconds: 5
    # Container image for the Storage service.
    image: vesoft/nebula-graphd
    logVolumeClaim:
      resources:
        requests:
          storage: 2Gi
      storageClassName: local-sc
    dataVolumeClaims:
    - resources:
        requests:
          storage: 2Gi
      storageClassName: local-sc
    replicas: 1
    resources:
      limits:
        cpu: "1"
        memory: 1Gi
      requests:
        cpu: 500m
        memory: 500Mi
    version: v3.6.0
    # Custom flags configuration for the Storage service.
    config: {}

For more detailed information about these parameters, see the Cluster configuration parameters section below.

Create the NebulaGraph cluster.
```
kubectl create -f nebulacluster.yaml -n nebula
```
Output:
```
nebulacluster.apps.nebula-graph.io/nebula created
```
If you don't specify the namespace using -n, it will default to the default namespace.

Check the status of the NebulaGraph cluster.

kubectl get nebulaclusters nebula -n nebula

Output:

NAME      READY   GRAPHD-DESIRED   GRAPHD-READY   METAD-DESIRED   METAD-READY   STORAGED-DESIRED   STORAGED-READY   AGE
nebula    True    1                1              1               1             1                  1                86s

Use `helm`¶

Add the NebulaGraph Operator Helm repository (if it's already added, run the next step directly).
```
helm repo add nebula-operator https://vesoft-inc.github.io/nebula-operator/charts
```
Update the Helm repository to fetch the latest resources.
```
helm repo update nebula-operator
```

Set environment variables for the configuration parameters required for installing the cluster.

export NEBULA_CLUSTER_NAME=nebula         # Name of the NebulaGraph cluster.
export NEBULA_CLUSTER_NAMESPACE=nebula    # Namespace for the NebulaGraph cluster.
export STORAGE_CLASS_NAME=local-sc       # StorageClass for the NebulaGraph cluster.

Create a namespace for the NebulaGraph cluster if it is not created.
```
kubectl create namespace "${NEBULA_CLUSTER_NAMESPACE}"
```
Check the customizable configuration parameters for the nebula-cluster Helm chart of the nebula-operator when creating the cluster.
- Run the following command to view all the configurable parameters.
```
helm show values nebula-operator/nebula-cluster
```
- Visit nebula-cluster/values.yaml to see all the configuration parameters for the NebulaGraph cluster. Click on Chart parameters to see the parameter descriptions and their default values.

Create the NebulaGraph cluster.

You can use the --set flag to customize the default values of the NebulaGraph cluster configuration. For example, --set nebula.storaged.replicas=3 sets the number of replicas for the Storage service to 3.

helm install "${NEBULA_CLUSTER_NAME}" nebula-operator/nebula-cluster \ 
    # Specify the version of the cluster chart. If not specified, it will install the latest version by default.
    # You can check all chart versions by running the command: helm search repo -l nebula-operator/nebula-cluster
    --version=1.7.3 \
    # Specify the namespace for the NebulaGraph cluster.
    --namespace="${NEBULA_CLUSTER_NAMESPACE}" \
    # Customize the cluster name.
    --set nameOverride="${NEBULA_CLUSTER_NAME}" \
    --set nebula.storageClassName="${STORAGE_CLASS_NAME}" \
    # Specify the version for the NebulaGraph cluster.
    --set nebula.version=v3.6.0

Check the status of NebulaGraph cluster pods.

kubectl -n "${NEBULA_CLUSTER_NAMESPACE}" get pod -l "app.kubernetes.io/cluster=${NEBULA_CLUSTER_NAME}"

Output:

NAME                               READY   STATUS    RESTARTS   AGE
nebula-exporter-854c76989c-mp725   1/1     Running   0          14h
nebula-graphd-0                    1/1     Running   0          14h
nebula-graphd-1                    1/1     Running   0          14h
nebula-metad-0                     1/1     Running   0          14h
nebula-metad-1                     1/1     Running   0          14h
nebula-metad-2                     1/1     Running   0          14h
nebula-storaged-0                  1/1     Running   0          14h
nebula-storaged-1                  1/1     Running   0          14h
nebula-storaged-2                  1/1     Running   0          14h

Cluster configuration parameters¶

The table below lists the configurable parameters and their descriptions for creating a NebulaGraph cluster using a YAML file.

Parameter	Default Value	Description
`metadata.name`	-	The name of the created NebulaGraph cluster.
`spec.console`	-	Launches a Console container for connecting to the Graph service. For configuration details, see nebula-console.
`spec.topologySpreadConstraints`	-	Controls the scheduling strategy for Pods. For more details, see Topology Spread Constraints. When the value of `topologyKey` is `kubernetes.io/zone`, the value of `whenUnsatisfiable` must be set to `DoNotSchedule`, and the value of `spec.schedulerName` should be `nebula-scheduler`.
`spec.graphd.replicas`	`1`	The number of replicas for the Graphd service.
`spec.graphd.image`	`vesoft/nebula-graphd`	The container image for the Graphd service.
`spec.graphd.version`	`v3.6.0`	The version of the Graphd service.
`spec.graphd.service`		Configuration for accessing the Graphd service via a Service.
`spec.graphd.logVolumeClaim.storageClassName`	-	The storage class name for the log volume claim of the Graphd service. When using sample configuration, replace it with the name of the pre-created storage class. See Storage Classes for creating a storage class.
`spec.metad.replicas`	`1`	The number of replicas for the Metad service.
`spec.metad.image`	`vesoft/nebula-metad`	The container image for the Metad service.
`spec.metad.version`	`v3.6.0`	The version of the Metad service.
`spec.metad.dataVolumeClaim.storageClassName`	-	Storage configuration for the data disk of the Metad service. When using sample configuration, replace it with the name of the pre-created storage class. See Storage Classes for creating a storage class.
`spec.metad.logVolumeClaim.storageClassName`	-	Storage configuration for the log disk of the Metad service. When using sample configuration, replace it with the name of the pre-created storage class. See Storage Classes for creating a storage class.
`spec.storaged.replicas`	`3`	The number of replicas for the Storaged service.
`spec.storaged.image`	`vesoft/nebula-storaged`	The container image for the Storaged service.
`spec.storaged.version`	`v3.6.0`	The version of the Storaged service.
`spec.storaged.dataVolumeClaims.resources.requests.storage`	-	The storage size for the data disk of the Storaged service. You can specify multiple data disks. When specifying multiple data disks, the paths are like `/usr/local/nebula/data1`, `/usr/local/nebula/data2`, and so on.
`spec.storaged.dataVolumeClaims.storageClassName`	-	Storage configuration for the data disks of the Storaged service. When using sample configuration, replace it with the name of the pre-created storage class. See Storage Classes for creating a storage class.
`spec.storaged.logVolumeClaim.storageClassName`	-	Storage configuration for the log disk of the Storaged service. When using sample configuration, replace it with the name of the pre-created storage class. See Storage Classes for creating a storage class.
`spec.<metad\|storaged\|graphd>.securityContext`	`{}`	Defines the permission and access control for the cluster containers to control access and execution of container operations. For details, see SecurityContext.
`spec.agent`	`{}`	Configuration for the Agent service used for backup and recovery, and log cleaning functions. If you don't customize this configuration, the default configuration is used.
`spec.reference.name`	`{}`	The name of the controller it depends on.
`spec.schedulerName`	`default-scheduler`	The name of the scheduler.
`spec.imagePullPolicy`	`Always`	The image pull policy for NebulaGraph images. For more details on pull policies, please see Image pull policy.
`spec.logRotate`	`{}`	Log rotation configuration. For details, see Managing Cluster Logs.
`spec.enablePVReclaim`	`false`	Defines whether to automatically delete PVCs after deleting the cluster to release data. For details, see Reclaim PV.

Last update: November 28, 2023