Deploy Data Source Connector

Supported Sources

You can deploy Data Source Connectors for the following sources:

• Physical Server
• Microsoft SQL Server
• Oracle Server
• Kubernetes/OpenShift

Data Source Connector System Prerequisites

Make sure that the Data Source Connector VM that you deploy for your Data Source Connection meets the following system requirements:

• 4 CPUs
• 16 GB RAM
• 171 GB disk space
• Outbound Internet connection

Port Requirements

Ensure that the following ports are open to allow communication between the Data Source Connector(s) and the data sources.

Physical Servers

Microsoft SQL Servers

Oracle Servers

Ensure that the following ports are open to allow communication between the Data Source Connector(s) and Oracle Server:

IBM Backup Service

Ensure that the following ports are open to allow communication between one or more Data Source Connectors and IBM backup service, as well as the IBM Cloud Storage.

Others

DataNetwork and SecondaryNetwork

• To deploy the Data Source Connector on a dual network, select different VLANs in DataNetwork and SecondaryNetwork, respectively.

• The Data Source Connector must have dual IP addresses if your data sources are in a private non-routable VLAN.

• Once you have deployed the Data Source Connector on a single network, you cannot modify the Data Source Connector to use a dual network or vice versa.

  10. On the Customize template page, enter the network settings: Network IP Address, Network Netmask, and Default Gateway. If you have selected a different VLAN for the secondary network, enter the Network IP Address, Network Netmask, and Default Gateway for the secondary network, as well. Click Next.

• To set the network settings using static IP addresses, manually enter the details in the respective fields for both DataNetwork and SecondaryNetwork.

• To set the network settings using DHCP, leave the fields blank in both the DataNetwork and SecondaryNetwork sections.

• Data Network and Secondary Network must be configured using the same network configuration method. That is static IP addresses or DHCP.

  11. Review the summary on the Ready to complete page and click Finish.

  12. Once the VM is created, power it on. After it boots, the services in the Data Source Connector VM (including the UI) can take 4-5 minutes to start.

5. Enter the IP address of the Data Source Connector VM in the address bar of your browser and click Enter.

6. On the Data Source Connector's User Interface, enter admin in the Username and Password fields to log in to the Data Source Connector.

On the next screen, you are prompted to change your password. Change your default password and log in again with your new password.

7. Verify the network configuration settings, make necessary changes, and click Continue.

8. On the Data Source Connector Configuration page, paste the Connection token in the Connection Claim Token field and click Save.

It can take another few minutes for the Data Source Connector to authenticate to the IBM Cloud Backup and Recovery Service. Click Data Source Connection to list the Data Source Connector(s) that are claimed.

VPE Gateways

While the VSI can connect to the Backup and Recovery instance, a VPE gateway provides a better performance, To create a VPE gateway, follow these steps.

1. From the IBM Cloud catalog, search for Virtual private endpoint and click Virtual private endpoint for VPC tile.
2. Select the appropriate Location, Name, Resource Group, VPC.
3. Under Request connection to a service, Select IBM Cloud Service.
4. Select Cloud service offerings to be Backup and Recovery and the appropriate Cloud service regions.
5. Select the appropriate Backup and Service instance in the table below.
6. Configure the Reserved IP as required and create the VPE Gateway.

Resource requirements

Ensure that your cluster has sufficient CPU and memory resources. The Data Source Connector will consume resources from your cluster nodes. Additional backup agent components (Datamover and Velero) will be deployed later during source registration.

Data Source Connector resource consumption:

Note: Datamover (DaemonSet) and Velero components are deployed automatically during source registration and are not part of the initial Data Source Connector installation.

1. Open IBM Cloud Shell (or use your local workspace if your cluster has a public endpoint).

2. Identify the source cluster where you want to install the Data Source Connector. This should be the Kubernetes or OpenShift cluster that you want to back up and protect. You must have admin access to this cluster to install the Data Source Connector.

   List the available clusters:

   ibmcloud ks cluster ls
   

   From the output, note the cluster name where the Data Source Connector should be deployed. Ensure you have admin privileges for this cluster.

3. Download and configure the KUBECONFIG for the selected cluster with admin privileges:

   ibmcloud ks cluster config --cluster <cluster-name> --admin
   

4. The Helm chart is hosted in the IBM Container Registry (ICR). Log in to the Helm/OCI registry by using the following command:

   helm registry login icr.io --username iamapikey --password "${API_KEY}"
   

   See Creating an API_KEY to create a new API_KEY if you don't have an existing one.

5. Retrieve the Helm install command that you copied earlier in the Create a data source connection section and update it based on your cluster type.

   For IKS Classic clusters:

   You must specify a storage class that is available on Classic clusters and disable SCC (Security Context Constraints) as it's specific to OpenShift:

   helm upgrade --install <k8-app-name> oci://icr.io/ext/brs/brs-ds-connector-chart --version 7.2.18-release-20260226-49768040 --set secrets.registrationToken=<your-registration-token> --set deploymentPlatform.rocp.sccEnabled=false --set volumeClaimTemplate.storageClass=ibmc-block-bronze --namespace ibm-brs-data-source-connector --create-namespace
   

   For IKS VPC clusters:

   The default storage class ibmc-vpc-block-metro-5iops-tier is used automatically. Disable SCC as it's specific to OpenShift:

   helm upgrade --install <k8-app-name> oci://icr.io/ext/brs/brs-ds-connector-chart --version 7.2.18-release-20260226-49768040 --set secrets.registrationToken=<your-registration-token> --set deploymentPlatform.rocp.sccEnabled=false --namespace ibm-brs-data-source-connector --create-namespace
   

   For ROKS Classic clusters:

   You must specify a storage class that is available on Classic clusters. SCC is enabled by default for OpenShift:

   helm upgrade --install <k8-app-name> oci://icr.io/ext/brs/brs-ds-connector-chart --version 7.2.18-release-20260226-49768040 --set secrets.registrationToken=<your-registration-token> --set volumeClaimTemplate.storageClass=ibmc-block-bronze --namespace ibm-brs-data-source-connector --create-namespace
   

   For ROKS VPC clusters:

   The default storage class ibmc-vpc-block-metro-5iops-tier is used automatically. SCC is enabled by default for OpenShift:

   helm upgrade --install <k8-app-name> oci://icr.io/ext/brs/brs-ds-connector-chart --version 7.2.18-release-20260226-49768040 --set secrets.registrationToken=<your-registration-token> --namespace ibm-brs-data-source-connector --create-namespace
   

   The default storage class ibmc-vpc-block-metro-5iops-tier is only available on VPC clusters and will not work on Classic clusters. For Classic clusters, use ibmc-block-bronze, ibmc-block-silver, or ibmc-block-gold.

6. Run the appropriate Helm install command in the IBM Cloud Shell based on your cluster type.

7. Check that the Helm release is installed:

   helm list -n ibm-brs-data-source-connector
   

   Check the data source connector pods status:

   kubectl get pods -n ibm-brs-data-source-connector
   

   Customizing the Helm install command

   You can customize the Helm install command with the following optional flags:

   • --namespace <namespace-name>: Specifies the namespace where the Data Source Connector will be deployed. If not specified, the default namespace is used.

   • --create-namespace: Creates the specified namespace if it doesn't exist. This flag is useful when deploying to a new namespace.

   • --set volumeClaimTemplate.storageClass=<storage-class-name>: Specifies the StorageClass to use for provisioning the persistent volume for the Data Source Connector. This parameter sets the storage class specifically for the Data Source Connector's persistent volume. This is required for Classic clusters as the default VPC storage class (ibmc-vpc-block-metro-5iops-tier) cannot be used on Classic clusters. If you want to use VPC File CSI driver storage classes, you need to create a custom storage class with GID:1000 and UID:1000 because the Data Source Connector runs as non-root app. For reference on creating storage classes, see Deploying an app that runs as non-root.

   • --set replicaCount=<number>: Sets the number of Data Source Connector pod replicas. The default value is 3. Adjust this based on your high availability and workload requirements.

   • --set fullnameOverride=<name>: Assigns a specific name to the Data Source Connector pods, making them easier to identify and manage.

   • --set deploymentPlatform.rocp.sccEnabled=false: Disables Security Context Constraints (SCC) for the deployment. This flag should be used for IBM Kubernetes Service clusters, as SCC is specific to OpenShift clusters.

   • --set nodeSelector.<key>="<value>": Schedules Data Source Connector pods on nodes with the specified label. This is useful when you want to run the connector on dedicated worker nodes. Use the default worker pool label ibm-cloud.kubernetes.io/worker-pool-name=<worker-pool-name> to target a specific worker pool.

   • --set "tolerations[<index>].key=<key>": Sets the toleration key for node taints. Tolerations allow pods to be scheduled on nodes with matching taints.

   • --set "tolerations[<index>].operator=<operator>": Specifies the toleration operator. Common values are Equal (exact match) or Exists (key exists).

   • --set "tolerations[<index>].value=<value>": Sets the toleration value that must match the taint value on the node.

   • --set "tolerations[<index>].effect=<effect>": Specifies the taint effect. Common values are NoSchedule (prevents scheduling), PreferNoSchedule (tries to avoid scheduling), or NoExecute (evicts existing pods).

   Adding a dedicated worker pool for Data Source Connector

   It's recommended to create a dedicated worker pool for the Data Source Connector. This helps ensure that the connector pods run on a dedicated worker pool with appropriate taints to ensure workload isolation and performance.

   1. Deploy a dedicated worker pool. All worker pools automatically have the label ibm-cloud.kubernetes.io/worker-pool-name=<worker-pool-name> which can be used for node selection.

   2. Add taints to the worker pool:

      ibmcloud ks worker-pool taint set \
          --cluster "${CLUSTER_NAME_ID}" \
          --worker-pool "${WORKER_POOL_NAME}" \
          --taint "dedicated=data-source-connector:NoSchedule" -f
      

   3. Deploy the Data Source Connector to the dedicated worker pool.

   Example Helm install command:

   helm upgrade --install dsc-test oci://icr.io/ext/brs/brs-ds-connector-chart \
      --version 7.2.18-release-20260226-49768040 \
      --namespace ibm-brs-data-source-connector \
      --create-namespace \
      --set secrets.registrationToken=xxx \
      --set fullnameOverride=dsc \
      --set replicaCount=2 \
      --set nodeSelector.ibm-cloud\\.kubernetes\\.io/worker-pool-name="my-worker-pool" \
      --set "tolerations[0].key=dedicated" \
      --set "tolerations[0].operator=Equal" \
      --set "tolerations[0].value=data-source-connector" \
      --set "tolerations[0].effect=NoSchedule"
   

   Replace my-worker-pool with your actual worker pool name.

Check your current Data Source Connector version

To find the currently installed version of the Data Source Connector:

helm list -n ibm-brs-data-source-connector

This command displays the installed Helm releases in the ibm-brs-data-source-connector namespace, including the chart version (APP VERSION column).

Alternatively, you can check the version using kubectl:

kubectl get statefulset -n ibm-brs-data-source-connector -o jsonpath='{.items[0].spec.template.spec.containers[0].image}'

This command shows the container image tag, which includes the version number.

Check for available versions

To view all available versions of the Data Source Connector chart, first log in to your IBM Cloud account:

ibmcloud login -a cloud.ibm.com --apikey "${API_KEY}"

Then, log in to the IBM Container Registry:

ibmcloud cr login

After logging in, list the available Data Source Connector images:

ibmcloud cr images --restrict ext/brs/brs-ds-connector

This command displays all available Data Source Connector versions. Compare the output with your currently installed version to determine if an upgrade is available.

Upgrade to a newer version

Once you've identified the target version, use the Helm upgrade command. You must provide the registration token because the old token may have expired.

helm upgrade --install <release-name> oci://icr.io/ext/brs/brs-ds-connector-chart --version <new-version> --reuse-values --set secrets.registrationToken=<token> -n ibm-brs-data-source-connector

Replace <release-name> with your actual Helm release name (use helm list -n ibm-brs-data-source-connector to find it), <new-version> with the target version number, and <token> with the registration token obtained from the Data Source Connection.

To obtain the registration token:

1. Log in to your IBM Cloud Backup and Recovery instance in the IBM Cloud console.
2. Navigate to System > Data Source Connections.
3. Locate the Data Source Connection that corresponds to your cluster.
4. Click on the three dots (⋮) menu corresponding to the connection.
5. Select Connection Token from the menu options.
6. Copy the displayed connection token (also referred to as the registration token).
7. Use this token in the secrets.registrationToken parameter of the upgrade command.

You must provide the registration token in the upgrade command because the previous token may have expired.

If you need to create a new Data Source Connection, see Create a data source connection for Kubernetes/OpenShift.

The --reuse-values flag preserves your existing configuration settings during the upgrade.

Example upgrade command:

helm upgrade --install dsc-test oci://icr.io/ext/brs/brs-ds-connector-chart --version 7.2.19-release-20260301-12345678 --reuse-values --set secrets.registrationToken=<new-token> -n ibm-brs-data-source-connector

Verify the upgrade

After upgrading, verify that the new version is running:

helm list -n ibm-brs-data-source-connector

Check the pod status to ensure all pods are running with the new version:

kubectl get pods -n ibm-brs-data-source-connector

Alternative: Using a values.yaml file

As an alternative to using multiple --set flags, you can use a values.yaml file to pass configuration values. This approach is cleaner and easier to manage, especially when dealing with complex configurations.

1. Create a custom-values.yaml file with the following content:

   secrets:
     registrationToken: "<your-registration-token>"
   
   nodeSelector:
     ibm-cloud.kubernetes.io/worker-pool-name: "<your-worker-pool-name>"
   
   tolerations:
     - key: "dedicated"
       operator: "Equal"
       value: "data-source-connector"
       effect: "NoSchedule"
   

2. Run the Helm install command using the values file:

   helm upgrade --install dsc oci://icr.io/ext/brs/brs-ds-connector-chart \
     --version 7.2.18-release-20260226-49768040 \
     --namespace ibm-brs-data-source-connector \
     --create-namespace \
     -f custom-values.yaml