Service jobs in Amazon Batch
Amazon Batch service jobs enable you to submit requests to Amazon services through Amazon Batch job queues. Currently, Amazon Batch supports SageMaker Training jobs as service jobs. Unlike containerized jobs where Amazon Batch manages the underlying container execution, service jobs allow Amazon Batch to provide job scheduling and queuing capabilities while the target Amazon service (such as SageMaker AI) handles the actual job execution.
Amazon Batch for SageMaker Training jobs allows data scientists to submit training jobs with priorities to configurable queues, ensuring workloads run without intervention as soon as resources are available. This capability addresses common challenges such as resource coordination, preventing accidental overspending, meeting budget constraints, optimizing costs with reserved instances, and eliminating the need for manual coordination between team members.
Service jobs differ from containerized jobs in several key ways:
-
Job submission: Service jobs must be submitted using the SubmitServiceJob API. Service jobs cannot be submitted through the Amazon Batch console.
-
Job execution: Amazon Batch schedules and queues service jobs, but the target Amazon service runs the actual job workload.
-
Resource identifiers: Service jobs use ARNs that contain "service-job" instead of "job" to distinguish them from containerized jobs.
To get started with Amazon Batch service jobs for SageMaker Training, see Getting started with Amazon Batch on SageMaker AI.