Run lifecycle in a HealthOmics workflow - AWS HealthOmics
Run status valuesTask RetriesPricing implications of run status

Run lifecycle in a HealthOmics workflow

You can track the progress of a run by monitoring the run status. HealthOmics updates the run status as a run proceeds through its lifecycle.

You can retrieve the current status of a run using the GetRun API operation. The HealthOmics console displays the status of each run on the Runs page.

Run status values

When you start a run, HealthOmics sets the run status to Pending. As the run proceeds through its lifecycle, HealthOmics updates the status value to reflect its current progress.

Note

You don't incur charges during any run status other than Running. See the next section for details.

HealthOmics supports the following run status values:

Pending

The run is in the queue, waiting to start. Runs typically remain in Pending for a brief period before they start.

  • Runs can remain in Pending for a longer time if you submit many jobs at the same time.

  • Runs remain in Pending after your account reaches the maximum number of concurrent runs.

  • A run remains in Pending if the run is part of a run group that has reached any of its resource maximum values.

  • You can adjust run priorities so that specific queued runs start before others. For more information about run priority, see Run priority.

Starting

HealthOmics creates the run and provisions the resources required for the run (such as temporary run storage and the engine node).

  • HealthOmics provisions temporary run storage at the start of the run, and deprovisions the run storage when the run is Stopping.

Running

A run remains in Running status during the import process, the processing of each task, and the export process.

  • HealthOmics imports the input files to the temporary run storage file system. The input files are read-only, to prevent tasks from modifying the inputs to other tasks in a workflow.

  • During file export, HealthOmics exports the output files from the run storage file system to the S3 location.

  • HealthOmics delivers the run logs and task logs to CloudWatch in real time while the run status is Running. For more information, see Logs in CloudWatch .

Stopping

After completion of the export process, the run transitions to the Stopping status.

  • HealthOmics deprovisions all resources (including the run storage file system and the engine node).

Completed

The run transitions to Completed after HealthOmics completes the resource deprovisioning.

  • HealthOmics has completed all run tasks and exported the output data without error.

  • The run outputs are available in the specified Amazon S3 URI output location and final run manifest logs and engine logs (if applicable) are available in CloudWatch.

  • For runs that support task retries, a run with Completed status can include one or more tasks that failed. As long as a task retry succeeded for each failed task, HealthOmics transitions the run to Completed. HealthOmics assigns a new task ID to each retry, so the run includes task IDs for the failed attempts and the completed attempt.

Failed

HealthOmics encountered one or more errors and failed to complete all the run tasks.

  • A failed run transitions through Stopping status while HealthOmics deprovisions the resources.

Canceled

A user initiated a request to cancel the run.

  • HealthOmics stops any running tasks and deprovisions all resources.

  • HealthOmics doesn't export any run output data when a user cancels a run. You don't have access to any intermediate files for a canceled run.

  • Your account incurs charges for the tasks and resources that the run consumed during Running status before the cancellation.

  • There are no charges if you cancel a run in Pending or Starting status.

Task Retries

If a task fails during a run, HealthOmics retries the task again, in the following situations:

  • For a WDL workflow, HealthOmics supports task retry when the task failed because of service errors (5XX HTTP status codes).

    By default, HealthOmics attempts up to two retries of a failed task. You can opt out of task retries by configuring the WDL definition file. For an example configuration, see Task resources in a HealthOmics workflow definition.

  • For a Nextflow workflow, you can configure retry conditions for tasks in the workflow defnition.

  • If every task in the run eventually completes, even if they required retries, HealthOmics transitions the run to Completed.

  • HealthOmics assigns a new task ID to each retry, so the run includes task IDs for the failed attempts and the completed attempt.

Pricing implications of run status

Your account can incur charges while the run status is Running. You don't incur charges during any other run status. For example, there is no charge for resources when the run is Starting or Stopping.

A run with Running status has the following billing implications:

  • Your account incurs charges for run storage file-system usage while the run status is Running. For information about the run storage types, See Run storage types in HealthOmics workflows.

  • Your account incurs charges for running tasks, based on the compute and memory resources that you specified for each task in the workflow definition, and based on the task duration. For more information, see Compute and memory requirements for HealthOmics tasks.

  • Each task has a minimum billing threshold of one minute. If you run a task for less than a minute, you incur a charge for the minimum one minute of usage. If possible, group small tasks together to optimize costs. Grouping tasks also reduces run time by avoiding the spin-up of multiple sequential tasks.

For additional information about HealthOmics pricing, see the HealthOmics Pricing.