Manage Compute

This section covers the basic functions of creating, changing, or removing compute clusters in your AI Data Platform Workbench.

Topics:

About Compute Clusters

All-purpose compute clusters provide you the compute resources to process your workloads in an AI Data Platform Workbench instance.

You manage your compute clusters from the Compute page in your Oracle AI Data Platform Workbench.


AI Data Platform Compute page with Compute highlighted in left pane

Types of Compute

Two types of compute exist in your AI Data Platform Workbench: all-purpose compute clusters and Default Master Catalog Compute Cluster.

You can only create all-purpose compute clusters in your AI Data Platform Workbench. All-purpose compute clusters are suitable for a versatile range of workloads and can be attached to your notebooks and used in workflows. Unless otherwise specified, any references to 'compute cluster' or 'cluster' in documentation refer to all-purpose compute clusters.

When you create a new all purpose compute cluster, you can choose either the Quickstart or Custom configuration. The Quickstart configuration is optimized to provide fast startup, while Custom configuration allows you to fine-tune your all purpose compute cluster to suit the specific workloads you need it to process. In both Quickstart and Custom configuration options, you can view cost projections and modify idle timeout options.

Note:

Installing custom libraries to a Quickstart configured all purpose compute cluster automatically changes it to the Custom configuration. This can impact startup performance.

Default Master Catalog Compute Cluster is present in all AI Data Platform Workbench instances. This cluster is responsible for essential AI Data Platform Workbench functions, like search crawls, refreshing catalog objects, creating, editing, and deleting objects, and testing connections.

Cluster Runtime

All-purpose compute clusters can be created with an Apache Spark 3.5 runtime. The runtime environment is compatible with:

  • Spark 3.5.0
  • Delta 3.2.0 (pre-included)
  • Python 3.11
  • Scala 2.12
  • Hadoop 3.3.4
  • Java 17

Compute Configuration Export and Import

Once you have created a compute cluster, you can export the settings you configured for that cluster to your workspace or a volume in a catalog to use again for other clusters. You can import your saved configurations to other created clusters or import them as part of the creation process.

Your configurations are available to other users with access to workspace or volume you saved the configuration to, making it easy to share configurations across teams.

Create a Quickstart Cluster

You can choose to create an all-purpose compute cluster with preconfigured settings to process data and AI workloads in your AI Data Platform Workbench.

The quickstart configuration is an Apache Spark cluster with 1 driver and up to 10 workers, each with AMD 2 OCPU and 32 GB memory. Autoscaling is enabled by default for quickstart configuration. You can set your clusters to be constantly active or you can set an interval of inactivity after which the cluster will automatically stop (idle timeout). Stopped clusters will resume when called on by an attached workflow or notebook.

You can edit your cluster at any time after creation.

  1. Click Create in the left navigation panel then click Compute. You can also navigate to your workspace and click Compute, then click Create cluster icon Create Cluster.

    Create cluster dialog open with Quickstart radial option selected

  2. Provide a name and description to identify your cluster.
  3. Select Runtime version.
  4. Select Quickstart as your cluster configuration.
  5. Select whether the number of workers is static or scales automatically. Autoscaling is enabled by default for quickstart configuration.
  6. For Run duration, select whether the cluster will stop running after a set duration of inactivity. If Idle timeout is selected, specify the idle time, in minutes, before the cluster will time out.
  7. Click Create.

Create a Custom Cluster

You can create an all-purpose compute cluster with configuration settings of your own choosing to process data and AI workloads in Oracle AI Data Platform Workbench.

Custom clusters are intended for advanced users who want to leverage the full range of configuration options to suit their needs. You should select the driver and worker options that best fit the workloads you are going to process. You can set your clusters to be constantly active or you can set an interval of inactivity after which the cluster will automatically stop (idle timeout). Stopped clusters will resume when called on by an attached workflow or notebook.

You can edit your cluster at any time after creation.

  1. Click Create in the left navigation panel then click Compute. You can also navigate to your workspace and click Compute, then click Create cluster icon Create Cluster.

    Create compute cluster dialog open with the Custom radial option selected

  2. For Create method, select Configure a new cluster.
  3. Provide a name and description to identify your cluster.
  4. Select Runtime version.
  5. Select Custom as your cluster configuration.
  6. Select the driver options for your cluster.
  7. Select the worker options for your cluster. These options apply to all cluster workers.
  8. Select whether the number of workers is static or scales automatically.
    • If Static amount, specify the number of workers.
    • If Autoscale, specify the minimum and maximum number of workers the cluster can scale to.
  9. For Run duration, select whether the cluster will stop running after a set duration of inactivity. If Idle timeout is selected, specify the idle time, in minutes, before the cluster will time out.
  10. Click Create.

Create a Cluster Using an Imported Configuration

You can create an all-purpose compute cluster with configuration settings exported from an existing compute cluster in your Oracle AI Data Platform Workbench.

Importing settings from an exported YAML file allows you to quickly replicate clusters with the same settings. You can import YAML files from AI Data Platform workspaces or volumes shared with you.

You can edit your cluster at any time after creation.

  1. Click Create in the left navigation panel then click Compute. You can also navigate to your workspace and click Compute, then click Create cluster icon Create Cluster.
  2. For Create method, select Use an exported configuration.
  3. From the Configuration file drop-down menu, select the exported configuration to use for your new cluster. Runtime, library, and environment settings are prefilled by the imported file.
  4. Provide a name and description to identify your cluster.
  5. Modify the remaining settings as needed.
  6. Click Create.

Create an NVIDIA GPU Cluster

You can choose to use an NVIDIA GPU in an All Purpose Compute Cluster to accelerate any workload in your unified AI and data pipeline.

NVIDIA GPU shapes use the following configurations:

Table 9-1 NVIDIA GPU Shapes

GPU Count OCPU Block storage (GB) GPU memory (GB) CPU memory (GB)
1 15 1500 24 240
2 30 3000 48 480

Note:

When you use NVIDIA GPU shapes, both the Driver and Worker shape must be an NVIDIA GPU. Mixing CPU and GPU shapes for the same cluster is currently not supported.
  1. Click Create in the left navigation panel then click Compute. You can also navigate to your workspace and click Compute, then click Create cluster icon Create Cluster.

    Create compute cluster dialog open with Custom radial option selected

  2. Provide a name and description to identify your cluster.
  3. Select Runtime version.
  4. Select Custom as your cluster configuration.
  5. For your cluster driver options:
    • Select NVIDIA GPU as the Driver Shape.
    • Select 1 or 2 as the GPU count.
  6. For your cluster worker options:
    • Select NVIDIA GPU as the Worker Shape.
    • Select 1 or 2 as the GPU count.
  7. Select whether the number of workers is static or scales automatically.
    • If Static amount, specify the number of workers.
    • If Autoscale, specify the minimum and maximum number of workers the cluster can scale to.
  8. For Run duration, select whether the cluster will stop running after a set duration of inactivity. If Idle timeout is selected, specify the idle time, in minutes, before the cluster will time out.
  9. Click Create.

Export Compute Configuration

You can export your compute cluster configuration and dependencies to a location in your workspace so it can be imported to other compute clusters.

  1. Navigate to your workspace and click Compute.
  2. Next to the compute cluster you want to export settings for, click Actions three dot icon Actions then click Export compute configuration. You can also click the cluster name, click Actions in the top-right, then click Export compute configuration.
  3. Verify the dependencies, such as libraries and environment variables. Deselect any that should not be exported.
  4. Choose a save location in your workspace or in a volume.
  5. Click Export.

Import Compute Configuration

You can import existing compute cluster configurations and their dependencies to an all-purpose cluster using YAML configuration files in your workspace.

  1. Navigate to your workspace and click Compute.
  2. Next to the compute cluster you want to import settings for, click Actions three dot icon Actions then click Import compute configuration. You can also click the cluster name, click Actions in the top-right, then click Import compute configuration.
  3. Search your workspace or volume for the configuration YAML files you want to upload and select one or more.
  4. Review the configuration files you selected then click Import.

Modify a Cluster

You can change settings or add additional parameters for your clusters.

  1. Navigate to your workspace and click Compute.
  2. Next to the compute cluster you want to modify, click Actions three dot icon Actions then click Edit.
  3. Modify the attributes of your compute cluster or add additional parameters as needed.
  4. Click Save.

Delete a Cluster

You can delete compute clusters that are unused or no longer needed.

  1. Navigate to your workspace and click Compute.
  2. Next to the cluster you want to delete, click Actions three dot icon Actions and click Delete.
  3. Click Delete.

View Cluster Details

You can review the shape and settings of a cluster at any time.

  1. Navigate to your workspace and click Compute.
  2. Click the name of the cluster you want to view details for.
  3. Click the Details tab.

Maintenance Updates for Compute Clusters

Oracle AI Data Platform compute automatically applies maintenance updates without user intervention.

The maintenance updates cover any necessary security patches or bug fixes for operating system and AI Data Platform internal components. AI Data Platform verifies there are no running clusters before applying these monthly maintenance updates.