Adding Nodes

This section describes how to add one or more nodes to an installed fabric by using the Data Fabric UI.

Adding nodes to an installed fabric can help in several ways. For example, you can add nodes to increase the total storage, redistribute resources, boost the compute capacity, or improve the availability of certain services. Adding nodes is sometimes referred to as a scaling operation because it allows you to scale the storage and performance of the fabric.

Limitations for Adding Nodes

Note the following limitations for adding nodes by using the Data Fabric UI:
  • The Data Fabric UI supports adding nodes to on-premises deployments only. Adding nodes to cloud deployments, such as AWS, Azure, and GCP, is currently not supported.
  • You can only add nodes to an as-a-service fabric and only if the fabric is running release 7.7.0 or later. You cannot use the Data Fabric UI to add nodes to a customer-managed cluster.

    To add nodes to a customer-managed cluster, you must use the Installer for customer-managed platforms or manual steps. See Extending a Cluster by Adding Nodes or Adding Nodes to a Cluster.

  • If the fabric has fewer than three nodes, adding nodes requires restarting ZooKeeper and Warden one node at a time on the previously installed nodes. During the restart phase, access to each node is lost temporarily. If the fabric has three or more nodes, the restart is not needed.
  • On any given fabric, only one scaling operation can be done at a time. If a scaling operation is already in progress, you cannot initiate another scaling operation on the same fabric. However, scaling operations can be performed at the same time on different fabrics belonging to the same global namespace.
  • Adding nodes is not possible when an upgrade operation is in progress. In addition, Upgrading a Data Fabric is not supported when an operation to add nodes is in progress.
  • The Data Fabric UI does not currently provide an option for removing a node from a fabric.

Prerequisites for Adding Nodes

Note the following prerequisites for adding nodes by using the Data Fabric UI:
  • To add nodes, you must have fabric manager credentials.
  • You can add nodes to any fabric, including a one-node fabric. There is no restriction on the size of a fabric for adding nodes.
  • Adding nodes to an on-premises installation requires that you provide the host nodes before starting the scaling operation. There is no restriction on the number of nodes you can add, but the nodes must meet certain requirements. Before beginning the operation, review Prerequisites for On-Premises Installation.

Steps for Adding Nodes

Use the following steps to add nodes:
  1. Log on to the Data Fabric UI.
  2. Select the Fabric manager option.
  3. On the Global namespace card , check the table view.
  4. Under the Resource Name column, click the name of the fabric for which you want to add nodes. The Fabric details are displayed.
  5. Click the Nodes tab. For more information, see Viewing Node Information.
  6. Click Add node. The Add node side drawer appears.
  7. Fill in the required node parameters:
    For this parameter . . . Do this
    SSH credentials* Enter your Username for SSH access to the node.

    Enter your Password.
    Nodes* Enter the fully qualified domain name (FQDN) of the first node you want to add. To add another node, click + Add node, and enter the next FQDN. If you enter a node name by mistake, you can remove it. Click the trash can () icon.
    Network settings If necessary, specify the optional network settings:
    EDF subnet – Specify one or more comma-separated subnet masks. For example:
    10.10.15.0/24,10.10.16.0/24
    EDF external – Specify a comma-separated list of tuples of host names and external IP addresses. For example:
    node1:1.1.1.1,node2:1.1.1.2,node3:1.1.1.3
    For more information about the network settings, see Using the MapR Subnet and MapR External Advanced Options.
    *This field is mandatory.
  8. Click Add node. The Data Fabric UI displays a Nodes in scale operation table that includes a Status column.
  9. Check the status of the scaling operation:
    To Do this
    View the status for a single node Hold your cursor over the information icon (). The icon is displayed when the node status is Installing. Hovering over the tooltip displays the installation status as a percentage.

    Or, in the Actions column, click the ellipsis for a node, and select View progress to display Node details information.
    View the status for all nodes Click View progress.
    If the scaling operation is successful, the Data Fabric UI displays a success message, and the table of active nodes is updated.

Troubleshooting the Scaling Operation

If the scaling operation fails, the Data Fabric UI displays a status of Failed in the Nodes in scale operation table. To obtain more information:
  1. Examine failure data for the entire scaling operation by clicking View progress above the table. Or, click the ellipsis () in the Action column, and select View progress (for a single node).
  2. If possible correct the root cause of the failure.
  3. Click Reinitiate to retry adding the node. The original side drawer is displayed, allowing you to make any needed configuration changes.
  4. Click Reinitiate.
If you are not able to resolve the issue that caused the scaling operation to fail, contact HPE Support.

Related maprcli Commands

To implement the features described on this page, the Data Fabric UI relies on the following maprcli commands. Links to these commands are provided for general reference. For more information, see maprcli Commands in This Guide.