StepConfig - Amazon EMR
ContentsSee Also

StepConfig

Specification for a cluster (job flow) step.

Contents

HadoopJarStep

The JAR file used for the step.

Type: HadoopJarStepConfig object

Required: Yes

Name

The name of the step.

Type: String

Length Constraints: Minimum length of 0. Maximum length of 256.

Pattern: [\u0020-\uD7FF\uE000-\uFFFD\uD800\uDC00-\uDBFF\uDFFF\r\n\t]*

Required: Yes

ActionOnFailure

The action to take when the step fails. Use one of the following values:

  • TERMINATE_CLUSTER - Shuts down the cluster.

  • CANCEL_AND_WAIT - Cancels any pending steps and returns the cluster to the WAITING state.

  • CONTINUE - Continues to the next step in the queue.

  • TERMINATE_JOB_FLOW - Shuts down the cluster. TERMINATE_JOB_FLOW is provided for backward compatibility. We recommend using TERMINATE_CLUSTER instead.

If a cluster's StepConcurrencyLevel is greater than 1, do not use AddJobFlowSteps to submit a step with this parameter set to CANCEL_AND_WAIT or TERMINATE_CLUSTER. The step is not submitted and the action fails with a message that the ActionOnFailure setting is not valid.

If you change a cluster's StepConcurrencyLevel to be greater than 1 while a step is running, the ActionOnFailure parameter may not behave as you expect. In this case, for a step that fails with this parameter set to CANCEL_AND_WAIT, pending steps and the running step are not canceled; for a step that fails with this parameter set to TERMINATE_CLUSTER, the cluster does not terminate.

Type: String

Valid Values: TERMINATE_JOB_FLOW | TERMINATE_CLUSTER | CANCEL_AND_WAIT | CONTINUE

Required: No

See Also

For more information about using this API in one of the language-specific AWS SDKs, see the following: