Help Center/ ModelArts/ User Guide for Senior AI Engineers (To Be Offline)/ Resource Pools (Old Version to Be Terminated)
Updated on 2023-09-06 GMT+08:00
View PDF
Share

Resource Pools (Old Version to Be Terminated)

ModelArts Resource Pools

New-version dedicated resource pools are now available in ModelArts. Experience the new pools. For more details, see Resource Pool Management.

This section describes only old-version dedicated resource pools, which will be discontinued soon. For details about the differences between the old and new dedicated resource pools, see Resource Pool.

When using ModelArts for full-process AI development, you can use two different resource pools.

  • Public Resource Pool: provides public large-scale computing clusters, which are allocated based on job parameter settings. Resources are isolated by job. You are charged based on resource specifications, service duration, and the number of instances, regardless of the job type (training job, deployment job, or development job). Public resource pools are provided by ModelArts by default. You do not need to create or configure them separately. During AI development, select a public resource pool.
  • Dedicated Resource Pool: provides exclusive compute resources, which can be used for notebook instances, training jobs, and model deployment. It delivers higher efficiency and cannot be shared with other users.

    Buy a dedicated resource pool and select the dedicated resource pool during AI development. For details about the dedicated resource pool, see the following:

    Dedicated Resource Pool

    Creating a Dedicated Resource Pool

    Scaling a Dedicated Resource Pool

    Deleting a Dedicated Resource Pool

Dedicated Resource Pool

  • Dedicated resource pools can be used by notebook instances, training jobs, TensorBoard, or for model deployment.
  • Dedicated resource pools are classified into two types: Dedicated for Development/Training and Dedicated for Service Deployment. The Dedicated for Development/Training type can be used only for notebook instances, TensorBoard, and training. The Dedicated for Service Deployment type can be used only for AI application deployment.
  • Dedicated resource pools are available only when they are in the Running state. If a dedicated resource pool is unavailable or abnormal, rectify the fault before using it.
  • After a dedicated resource pool is created, the billing starts based on the selected specifications.
  • Dedicated resource pools can be billed in pay-per-use or yearly/monthly mode.

Creating a Dedicated Resource Pool

  1. Log in to the ModelArts management console and choose Dedicated Resource Pools on the left.
  2. On the Dedicated Resource Pools page, select Dedicated for Development/Training or Dedicated for Service Deployment.
  3. Click Create in the upper left corner. The page for creating a dedicated resource pool is displayed.
  4. Set the parameters on the page. For details about how to set parameters, see Table 1 and Table 2.
    Table 1 Parameters of the Dedicated for Development/Training type

    Parameter

    Description

    Resource Type

    The default value is Dedicated for Development/Training and cannot be changed.

    Billing Mode

    Select a billing mode, Yearly/Monthly or Pay-per-use.

    Name

    Name of a dedicated resource pool.

    The value can contain letters, digits, hyphens (-), and underscores (_).

    Description

    Brief description of a dedicated resource pool.

    Nodes

    Select the number of nodes in a dedicated resource pool. More nodes mean higher computing performance and a higher cost.

    Specifications

    Required specifications. The GPU delivers better performance, and the CPU is more cost-effective. If a flavor is sold out, you can purchase it again only after other users delete the resource pool.

    Required Duration
    • Select the time length when you want to use the resource pool. This parameter is mandatory only when the Yearly/Monthly billing mode is selected.

      The duration ranges from one month to one year. ModelArts provides a 1-year preference package, which allows you to enjoy the product for 1 year by paying for only 10 months.

    • Auto renewal. With auto-renewal, the system automatically renews your product before the product expires.
    Table 2 Parameters of the Dedicated for Service Deployment type

    Parameter

    Description

    Resource Type

    The default value is Dedicated for Service Deployment and cannot be changed.

    Billing Mode

    Select a billing mode, Yearly/Monthly or Pay-per-use. The Yearly/Monthly billing mode is supported only in CN North-Beijing4.

    Name

    Name of a dedicated resource pool.

    Enter 4 to 24 characters starting with a lowercase letter and not ending with a hyphen (-). Only lowercase letters, digits, and hyphens (-) are allowed.

    Description

    Brief description of a dedicated resource pool.

    Custom Network Configuration

    If you enable Custom Network Configuration, the service instance runs on the specified network and can communicate with other cloud service resource instances on the network. If you do not enable Custom Network Configuration, ModelArts allocates a dedicated network to each user and isolates users from each other.

    If you enable Custom Network Configuration, set VPC, Subnet, and Security Group. If no network is available, go to the VPC service and create a network.

    AZ

    You can select Random, AZ 1, AZ 2, or AZ 3 based on site requirements. An AZ is a physical region where resources use independent power supplies and networks. AZs are physically isolated but interconnected through an internal network. To enhance workload availability, create nodes in different AZs.

    Nodes

    Select the number of nodes in a dedicated resource pool. More nodes mean higher computing performance and a higher cost.

    Specifications

    Required specifications. The GPU delivers better performance, and the CPU is more cost-effective.

    Required Duration
    • Select the time length when you want to use the resource pool. This parameter is mandatory only when the Yearly/Monthly billing mode is selected. The duration ranges from one month to 11 months.
    • Auto renewal. With auto-renewal, the system automatically renews your product before the product expires.
  5. After confirming that the specifications are correct, create a dedicated resource pool as prompted. After a dedicated resource pool is created, its status changes to Running.

    A newly purchased dedicated resource pool can be used only after its development environment is initialized.

(Optional) Interconnecting a VPC

After a resource pool is created, you can interconnect a VPC on the resource pool details page. The procedure is as follows:

Only resource pools dedicated for development/training can be interconnected with a VPC.

  1. On the Dedicated Resource Pools page, click the target resource pool.
  2. On the dedicated resource pool details page, click Configure NAS VPC.
    Figure 1 Configure NAS VPC
  3. On the Configure NAS VPC page, enable NAS VPC Connection and set NAS VPC and NAS Subnet.
    Figure 2 Configuring NAS VPC
    • If no VPC is available, click Create VPC on the right to create a VPC.
    • If no subnet is available, click Create Subnet on the right to create a subnet.
  4. Click OK.

Scaling a Dedicated Resource Pool

After a dedicated resource pool is used for a period of time, you can scale out or in the capacity of the resource pool by increasing or decreasing the number of nodes.

The procedure for scaling is as follows:

  1. Go to the dedicated resource pool management page, locate the row that contains the desired dedicated resource pool, and click Scale in the Operation column.
  2. On the scaling page, increase or decrease the number of nodes. Increasing the node quantity scales out the resource pool whereas decreasing the node quantity scales in the resource pool. Scale the capacity based on service requirements.
    • During capacity expansion, increase the number of nodes according to the quota of your account. If the number exceeds the quota, the capacity expansion will fail.
    • During capacity reduction, delete the target nodes in the Operation column. To reduce one node, switch off the node in Node List to delete the node. See Figure 3.

      Before reducing the capacity of a resource pool for real-time inference, ensure that there are no running instances on the nodes to be released. Otherwise, the real-time services will be interrupted. If you are not sure whether any instance is running on the node to be released, submit a consultation service ticket.

      Figure 3 Switching off a node to delete the node during scale-in
  3. Click Submit. After the request is submitted, the dedicated resource pool management page is displayed.

    The node is not deleted immediately after the request is submitted. In this case, do not delete nodes from the dedicated resource pool list. Otherwise, the deletion may fail.

    You can view the event list on the dedicated resource pool details page. "Begin to delete resource node %s" indicates that the node deletion starts. "Resource node %s deleted" indicates that the node has been deleted in the background.

Switching the Dedicated Resource Pool Billing Mode

  • Converting the billing mode of a dedicated resource pool from pay-per-use to yearly/monthly

    ModelArts allows you to change the billing mode of a dedicated resource pool to be trained from pay-per-use to yearly/monthly. The prerequisite is that use a pay-per-use dedicated resource pool for more than one consumption. If you do not have any consumption record of using the purchased pay-per-use dedicated resource pool, an error will be reported when you change the billing mode.

    To change the billing mode to yearly/monthly in the next subscription period, perform the following steps:

    1. Go to the dedicated resource pool management page, locate the row that contains the desired dedicated resource pool, and click Change Billing Mode in the Operation column. The page for changing the billing mode from pay-per-use to yearly/monthly is displayed.
    2. Select the yearly/monthly subscription duration as prompted and select whether to automatically renew the subscription. After the payment is complete, the conversion is successful.
  • Converting the billing mode of a dedicated resource pool from yearly/monthly to pay-per-use

    For a dedicated resource pool used for training with the billing mode of yearly/monthly, the billing mode can be set to pay-per-use after the yearly/monthly billing period expires.

    To change the billing mode to pay-per-use, perform the following steps:

    1. Log in to the ModelArts console and choose More > Billing Center > Renewal.
      Figure 4 Renewal
    2. On the Renewals page, locate the row that contains the dedicated resource pool, and click More and choose Change to Pay-per-Use After Expiration in the Operation column.

Deleting a Dedicated Resource Pool

If a dedicated resource pool is no longer needed for AI service development, you can delete the resource pool to release resources.

  • After a dedicated resource pool is deleted, the training jobs, notebook instances, real-time services, and batch services that depend on the resource pool will become unavailable. A dedicated resource pool cannot be restored after being deleted. Exercise caution when performing this operation.
  1. Go to the dedicated resource pool management page and release resources.
    • For a yearly/monthly dedicated resource pool, click Unsubscribe in the Operation column. After the unsubscription is complete, the resource is automatically deleted.
    • For a pay-per-use dedicated resource pool, click Delete in the Operation column.
    • For a dedicated resource pool failing to be created, click Delete in the Operation column to delete it.
  2. In the dialog box that is displayed, click OK.