Using UCP Expansion for UCP Converged Infrastructure VMware virtualization on an existing cluster

Unified Compute Platform (UCP) Advisor Administration Guide

Version
4.5.x
Audience
anonymous
Part Number
MK-92UCP119-13
ft:lastEdition
2024-03-11

Use UCP Expansion to add compute nodes to an existing cluster and onboard them to UCP Advisor.

Note:
  • UCP Expansion is not supported on HA840 G3 server model.
  • Distributed Resource Scheduler (DRS) and vSphere High Availability (HA) features are not managed.
  • Using UCP Expansion you can discover compute nodes, configure vCenter cluster, and onboard the compute nodes that are configured with FQDN.
  • vCenter 7.0.2 or later is preinstalled and configured, and the login credentials are set.
  • The compute nodes are configured with default or custom credentials.
  • The ESXi hosts used in the UCP expansion are not part of a vCenter cluster, vCenter inventory, or a VDS port group.
  • The disks on the compute nodes you want to discover are not partitioned.
  • A minimum of two Network Interface Controllers (NIC) are physically connected to the compute nodes.
  • At least one NIC is available for UCP Expansion configuration.
  • Hitachi custom UCP ESXi image is installed on all the compute nodes.
  • For ESXi raw image installed from the VMware website, you must enable the ESXi SOL settings. See Enabling SOL settings on ESXi server.
  • The SSH, SOL, and Shell services are not disabled.
  • The DNS and NTPD services are up and running on the ESXi server.
  • All nodes that you want to configure using UCP Expansion are in a single network.
  • The nodes selected for UCP Expansion and the nodes in the existing cluster are of the same model type.
  • The ESXi version of the nodes selected for UCP Expansion and the nodes in the existing cluster are at the same level.
  • The existing vCenter cluster and the ESXi servers are in Normal state, and connected to the vCenter.
  • If you want to use the FQDN configuration for the compute nodes, then verify that the FQDN address is reachable from UCP Advisor.
  • Configure NTP settings on HA800 series servers. See "Configuring NTP settings on HA800 series servers" in the Hitachi Unified Compute Platform (UCP) Software Installation Guide.

Basic Information

  1. In a web browser, log on to UCP Advisor.
  2. Click Settings > UCP Expansion.
    UCP Expansion page for CI flow

  3. For Select Solution Type, select UCP CI VMware virtualization.
  4. For the Search By option:
    • Select IPv4 Address if UCP Advisor and the compute nodes are in different subnets, and then for the IPv4 Address Range, enter the IPv4 address range for the compute nodes that you want to discover.
    • Select IPv6 Address if UCP Advisor and the compute nodes are in the same subnet.
  5. For the UCP Advisor Gateway IPv4 address, select the IPv4 address of the Master Node VM, which is the same IPv4 address that you used to log on to UCP Advisor, or select the external gateway VM IPv4 address from the list.

    Select Enter new IP Address to manually enter the gateway IPv4 address that you want to use for UCP Expansion configuration.

  6. (Optional) Enter the BMC username and password. The BMC credentials must be the same for all the compute nodes that you want to configure.
  7. (Optional) Enter the ESXi username and password. The ESXi credentials must be the same for all the compute nodes that you want to configure.
  8. Click Next.
  9. Enter the vCenter information:.
    vCenter Address
    Enter the vCenter IPv4 or the FQDN address.

    The vCenter version must be either at the same or higher level of operating system as the compute nodes that you want to add. You cannot add compute nodes that are at a higher level of operating system than the version used by vCenter.

    Username and Password
    Enter the vCenter username and password.
  10. Click Next to begin autodiscovery of the available and supported compute nodes. his might take several minutes according to the number of nodes.

Node Autodiscovery

  1. Select one or more compute nodes that you want to add to the cluster.
    Note:
    • You must select compute nodes that are of the same model type. For example, you can select any number of DS120 nodes; however, you cannot select a combination of DS120 nodes and DS220 nodes in a single operation.
    • Autodiscovery of HA810/HA820 G2 and HA810/HA820 G3 nodes takes approximately 25 minutes to complete.
    • Verify that the selected compute nodes have an operating system installed.
    • Compute nodes that are already assigned to a UCP system are not discovered during autodiscovery.
    • The compute nodes that you want to configure are not part of the vCenter.

    Compute nodes auto-discovered during UCP Expansion

    Note: If the compute nodes are listed with SOL errors, reset the BMC/iLO and clear any active IPMI sessions from the BMC/iLO console.
  2. Select one of the following options:
    • Click Configure Existing Cluster to add the selected nodes to an existing cluster.
    • Click Configure Devices to skip the vCenter cluster configuration and configure IPv4, BMC, and ESXi on the selected nodes, and then onboard them to a UCP system or Spare Pool.
    • Click Rediscover if all the compute nodes are not discovered or if the nodes are displayed with a warning message. Rectify the warning, and then rediscover the nodes.
    • Click Back to restart the UCP Expansion configuration.

IPv4 Assignment

  1. Select an option to assign IPv4 for the BMC.
    • Select the Server power on/up order option, enter Continue, and then click Submit to assign a sequence of IPv4 addresses to the selected compute nodes.
      Note: The compute nodes are first powered off. You must physically power on the compute nodes in a bottom-up sequence in the physical rack for the IPv4 assignment to succeed.

      Assign IPv4 for the discovered nodes during UCP Expansion.


      Server power on/up order

    • Select Manually Assign to manually assign IPv4 addresses to the selected compute nodes.

BMC Configuration

You can skip this step if the BMC is already configured for all of the selected compute nodes.

  1. Click Configure BMC.
    Configure BMC for UCP Expansion

  2. Enter the following network settings. When done, click Next, review the summary, and then click Configure.
    IPv4 Address Range
    Enter the IPv4 address range for the selected compute nodes. You can also enter individual IPv4 addresses separated by a comma.
    Default Gateway
    Enter the default gateway address for the BMC.
    Subnet Mask
    Enter the subnet mask IPv4 address for the BMC.
    VLAN ID
    (Optional) Enter the VLAN ID for the BMC.
    Note: You must enter the VLAN ID if you have configured it. If you enter an incorrectVLAN ID, the BMC will not be configured.
    NTP
    (Optional) Enter the NTP server IPv4 address for the BMC.
    BMC Admin User New Password
    Enter the BMC admin user password.
    BMC Admin User Confirm Password
    Confirm the BMC admin user password.
  3. Click Next to continue, or click Quit to skip the next steps and complete UCP expansion.
    Note: If BMC configuration fails, restart UCP Expansion. You cannot skip BMC configuration at this stage of UCP Expansion.

ESXi Configuration

You can skip this step if ESXi is already configured for all the selected compute nodes and if all the compute nodes are reachable.

  1. Click Configure ESXi.
    ESXi configuration for UCP Expansion

  2. Enter the following ESXi settings. When done, click Next, review the summary, and then click Configure.
    IPv4 Address Range
    Enter the IPv4 address range for the selected compute nodes. You can also enter individual IPv4 addresses separated by a comma.
    Default Gateway
    Enter the default gateway address for the ESXi server.
    Subnet Mask
    Enter the subnet mask IPv4 address for the ESXi server.
    NTP
    (Optional) Enter the NTP server IPv4 address for the ESXi server.
    DNS
    Enter the DNS IPv4 address.
    VLAN ID
    Enter the VLAN ID for the ESXi server.
    Note: You must enter the VLAN ID if you have configured it. If you enter an incorrect VLAN ID, the ESXi will not be configured.
    ESXi New Password
    Enter a new ESXi password for the root user.
    ESXi Confirm Password
    Confirm the ESXi root user password.
  3. Click Next to continue, or click Quit to skip the next steps and complete UCP expansion.
    Note: If ESXi configuration fails, restart UCP Expansion. You cannot skip ESXi configuration at this stage of UCP Expansion.

vCenter Cluster Configuration

  1. Enter the following vCenter settings.
    • Datacenter Details
      Datacenter Name
      Select a datacenter from the list.
      Cluster Name
      Select a cluster from the list.
    • Distributed Switch Configuration
      VDS Name
      Select a dvSwitch from the list.
      Max MTU
      Enter the maximum MTU value for the dvSwitch.
    • Management Portgroup Configuration
      Portgroup Name
      Select the management port group name.
    • vMotion Portgroup Configuration
      Portgroup Name
      Select the vMotion port group from the list.
      IPv4 Address Range
      Enter the IPv4 address range for the selected compute nodes. You can also enter individual IPv4 addresses separated by a comma.
      Default Gateway
      Enter the default gateway IPv4 address for the vMotion port group.
      Subnet Mask
      Enter the subnet mask IPv4 address for the vMotion port group.
      Max MTU
      Enter the maximum MTU value for the vMotion port group.
    • License Key
      ESXi License Key
      (Optional) Enter the ESXi license key.
  2. Click Next, review the settings, and then click Submit to configure vCenter cluster.
    vCenter cluster configuraton summary page

  3. Click Next to continue, or click Quit to skip the next steps and to complete UCP expansion without onboarding the compute devices.

Onboard Compute Devices

  1. Onboard the discovered compute nodes.
    Onboard the discovered compute nodes during UCP Expansion.

    • Click Onboard Devices, select the UCP system where you want to onboard the compute nodes, and then click Onboard.
      Note: You can onboard the compute nodes to a UCP system or to the Spare Pool.

      Onboard compute nodes to a UCP system.

      Click Finish to complete the UCP expansion.


      Compute nodes onboarded

      Note: If the onboarding for HA810/HA820 G2 or HA810/HA820 G3 compute nodes fails, verify that the BMC and the ESXi for that compute node are reachable, and then try onboarding them again.
    • Click Skip to skip onboarding of the compute nodes and to complete the UCP expansion.
    • Click Export as CSV to download the compute node details as a CSV file. You can later use this file to onboard the compute nodes to your UCP systems.

Complete UCP Expansion

After completing the UCP expansion:

  • Click UCP Expansion Log to download the log report and troubleshoot any issues.
    Note: You can also view the UCP expansion logs in the day0 folder of the UCP Advisor log bundle.
  • Click UCP Expansion Report to download the UCP expansion configuration report.
    UCP Expansion complete