Deploying a Feature Engineering Task - Teradata Vantage

ClearScape Analytics™ ModelOps User Guide

Deployment
VantageCloud
VantageCore
Edition
Enterprise
IntelliFlex
VMware
Product
Teradata Vantage
Release Number
7.1
Published
December 2024
ft:locale
en-US
ft:lastEdition
2024-12-13
dita:mapPath
zdn1704469623418.ditamap
dita:ditavalPath
azq1671041405318.ditaval
dita:id
rgn1654191066978
lifecycle
latest
Product Category
ClearScape
  1. From the Task - Lifecycle page, select Deploy.
    The Deploy feature engineering task dialog displays.
  2. In the Select Engine tab, set the following properties, then select Next.
    Property Description
    Engine type Specify the engine type for deployment:
    • Docker
    Deployment mode Specify the deployment mode for docker deployment:
    • Batch
  3. In the Publish tab, set the following properties, then select Next:
    Property   Description
    Connection Select the dataset connection to use when publishing the task. Teradata strongly recommends selecting a service connection.
    Publishing a task means building an artifact (docker image) with the code and libraries required to run the task.

    When a scheduled task runs, the published artifact is executed.

    Advanced options Engine name Select the Engine name from the drop-down for the deployment job.
    Docker image Select the Docker image from the drop-down for the deployment job.
  4. In the Schedule tab, set the following properties:
    Property   Description
    Enable scheduling Select the checkbox to enable scheduling.
    Scheduling options Specify scheduling options to run scoring job after time interval.
    • Radio button to choose one of the following: Cron, Manual, Run once, or Time scheduler.
    • Drop-down menu to choose if the schedule be Hourly, Daily, Weekly, Monthly, or Yearly.
    • Drop-down menu to set the interval to hours and minutes.
    Advanced options Resource template Select the resource template from the drop-down list.

    When Custom is selected, an additional option to provide allocated GPU resources appears.

    Build properties You have the option to provide key value pairs for it.
  5. Select Deploy.
    If you selected Run Once in the scheduling options, the job is run immediately.
  6. Close the dialog box when the deployment progress completes.
    The task status changes to Deployed. The Deployment details section displays on the page with the following details:
    Property Description
    Job ID Specifies the deployment job ID. You can select View Job Details to see the event details of the job.
    Date Specifies the deployment date.
    User Specifies the user name who executed the deployment job.
    Status Shows the status of the deployed job as Completed.
    Resources Specifies the resources utilized in the job, including CPU and Memory.
    Engine Specifies the deployment engine.
    Job progress Lists all phases of the deployment job:
    • Status: Status of each job event as Created or Completed
    • Start date/time: Start date and time for each job event
    • End date/time: End date and time for each job event
    • Duration: Duration of each job event