Task resources in a HealthOmics workflow definition

フォーカスモード

Task resources in a HealthOmics workflow definition - AWS HealthOmics

このページはお客様の言語に翻訳されていません。翻訳のリクエスト

In the workflow definition, define the following for each task:

The container image for task. For more information, see Container images in Amazon ECR for private workflows.
The number of CPUs and memory required for the task. For more information, see Compute and memory requirements for HealthOmics tasks.

HealthOmics ignores any per-task storage specifications. HealthOmics provides run storage that all tasks in the run can access. For more information, see Run storage types in HealthOmics workflows.

anchor anchor anchor


task my_task {
   runtime {
      container: "<aws-account-id>.dkr.ecr.<aws-region>.amazonaws.com/<image-name>"
      cpu: 2
      memory: "4 GB"
   }
   ...
}

For a WDL workflow, HealthOmics attempts up to two retries for a task that fails because of service errors (API request returns a 5XX HTTP status code). For more information about task retries, see Task Retries.

You can opt out of the retry behavior by specifying the following configuration for the task in the WDL definition file: