AWS HealthOmics

Workbench supports a number of engine parameters for AWS HealthOmics that can greatly impact the performance of your workflow runs. These parameters can be set when you create or update an engine in Workbench ( see Parameters) or when you submit a workflow run ( see Submitting and Monitoring a Workflow Run).

Available Parameters

storage_type=[DYNAMIC|STATIC]

The storage_type parameter allows you to specify the type of storage that will be used by the engine. The two options are DYNAMIC and STATIC.

• DYNAMIC: The engine will use a dynamic storage type. This means that the engine will automatically scale the storage based on the size of the data being processed. While you do not need to worry about running out of storage space, this may reduce performance and has lower IO

• STATIC: The engine will use a static storage type. This means that the engine will use a fixed amount of storage space associated with a Fsx Luster file store. This can improve performance and has higher IO, but you may run out of storage space if you exceed the limit.

storage_capacity=NUMBER

If you specify STATIC for the storage_type parameter, you must also specify the storage_size parameter. This will set the total size to allocate for the Fsx Luster file store in GiB.

log_level=[OFF|FATAL|ERROR|ALL]

The log_level parameter allows you to specify the level of logging that will be used by the engine and sent to cloud watch. The four options are OFF, FATAL, ERROR, and ALL.

• OFF: No logging will be performed by the engine

• FATAL: Only fatal errors encountered by the engine will be logged

• ERROR: All errors encountered by the engine will be logged

• ALL: All events will be logged by the engine

priority=NUMBER

The priority parameter allows you to specify the priority of the run with a number from 1-1000. The higher the number, the higher the priority. This can be useful if you have multiple runs that need to be processed to ensure more urgent runs are processed first.

retention_mode=[RETAIN|REMOVE]

The AWS HealthOmics API has quotas on the number of runs that can be stored. The retention_mode parameter allows you to specify whether a complete run should be kept or removed if and only if the quota is exceeded. The two options are RETAIN and REMOVE.

• RETAIN: The run will be kept even if the quota is exceeded. This may prevent you from submitting new runs

• REMOVE: The run will be removed if the quota is exceeded. This may allow you to submit new runs, but you will lose the metadata associated with the run. Run outputs stored in S3 will not be deleted.

run_group_id=STRING

In Aws HealthOmics, you can define a RunGroup to easily limit compute resources used across all runs within the run group. You can set the maximum vCPU, maximum duration, or maximum concurrent runs to help limit your use of compute resources. The run_group_id parameter allows you to specify the unique identifier for the run group.

accelerators=[GPU]

AWS HealthOmics supports attaching specialized hardware accelerators to your workflows and runs. These accelerators can have a significant impact on the performance of your workflows. The accelerators parameter allows you to specify the type of accelerator to attach to the engine. At this time, only option is GPU.