Updated on 2025-08-18 GMT+08:00

Expanding and Reducing Lite Server Supernodes

Scenario

You need to purchase supernodes to expand Lite Server resources due to service changes.

Regardless of whether the purchased supernodes and the existing supernodes belong to the same Xpod (physical supernode), hyperplane networks of the two vxpod supernodes are disconnected, and network communication is available only on the parameter plane. This section describes how to enable intra–hyperplane communication for the added supernodes.

Constraints

Scaling out

  • This applies only to the scale–out of Snt9b23 supernodes.
  • There cannot be more than 48 subnodes during the supernode scale–out.
  • Only available nodes can be scaled out.
  • The scale–out does not affect services on original nodes.

Scaling in

  • This applies only to the scale-in of Snt9b23 supernodes.
  • If the supernode has only one subnode, scale-in is not supported.
  • If the node to be scaled in is in an intermediate state, such as switching OS, restarting, starting, or stopping, scale-in is not supported.
  • Scaling in affects services on original nodes and the EVS disks of the nodes are released. Ensure that services have been migrated away from the nodes to be scaled in to prevent data loss.

Billing

Yearly/monthly is supported for both node scale-out and scale-in. A billing order will be generated for the new nodes.

Scaling out the Supernodes

  1. Log in to the ModelArts console.
  2. In the navigation pane on the left, choose Lite Servers under Resource Management. The Resources tab is displayed.
  3. Locate the target supernode in the list and choose > Change Supernode Specifications on the right.
  4. On the displayed page, view the resource information to be changed and select the target specifications. Sold-out resources are displayed in gray and cannot be selected.
  5. Set the parameters.
    Table 1 Parameters

    Parameter

    Description

    System Disk

    Select System Disk Type and set Size. A system disk is automatically created when you create a Lite Server. Set the system disk size to at least 100 GB. You cannot expand the system disk capacity on ECS after the Lite Server resource is created.

    Add Data Disk

    Click Add Data Disk to mount a data disk to the Lite Server. You can also mount data disks or expand the capacity of system disks on ECS after the Lite Server resource is created. For details, see Using EVS for Storage.

    Image

    • Public image

      Public images are available for all users. All users can read the image by image ID.

      ModelArts allows you to perform development and training directly without additional configuration as it provides multiple public images, supports multiple OSs, and has built–in AI drivers and software.

      For details about the supported public images, see Mapping Between Compute Resources and Image Versions.

    • Private image

      Only the image creator can use the image. You can select a private image to save your time from repeatedly configuring servers.

    Login Credential

    Key pair is recommended as it features higher security than Password. If you select Password, ensure that the password meets complexity requirements to prevent malicious attacks.

    • Key pair

      Use a key pair to log in to the Server node. You can select an existing key pair, or click Create Key Pair to create one.

      NOTE:

      If you use an existing key pair, ensure that you have saved the key file locally. Otherwise, logging in to the Server node will fail.

    (Optional) Custom Instance Injection

    Use this function to configure Server nodes if you want to:

    • Use scripts to simplify the Server node configuration.
    • Use scripts to initialize OSs.
    • Use existing scripts and upload them to the server when creating the Server node.
    • Use scripts for other purposes.

    Currently, As text and As file are supported. For details, see Injecting User Data into BMSs or Injecting User Data into ECSs.

  6. Click Next to preview the change information, view the node configuration and fee, and click Submit.
  7. After the scale-out is complete, expand the supernode that has been scaled out on the resource list page and check the running status of the new subnodes under the supernode.

Scaling in the Supernodes

  1. Log in to the ModelArts console.
  2. In the navigation pane on the left, choose Lite Servers under Resource Management. The Resources tab is displayed.
  3. Locate the target supernode in the list and choose > Change Supernode Specifications on the right.
  4. On the displayed page, view the resource information to be changed and select the target specifications. In the list of target reserved nodes, select the nodes to be reserved.

    Target specifications specify specifications of the nodes to be reserved.

    The number of reserved nodes must be the same as the number of compute nodes in the target specifications. Otherwise, the scale-in task fails to be submitted.

  5. Select I acknowledge the potential risks and agree to proceed with the scale-in and enter YES to confirm the scale-in.

    When scaling in, unreserved nodes and their associated EVS disks will be removed. Move your services off these nodes first to avoid losing data.

  6. Click Next to preview the change information, view the node reservation and deletion information, confirm the information, and click Submit.