easyMultiple Choice
SageMaker Pipelines Condition Step for Failure Handling
A company is using SageMaker Pipelines to automate a multi-step ML workflow. The pipeline includes data preprocessing, training, and model evaluation. The team wants to ensure that if the evaluation step fails, the pipeline stops and sends an alert to the operations team. Which SageMaker Pipelines feature should they use?
Quick Answer
Stopping a pipeline and alerting a team when a specific step's output indicates failure requires two things working together: a way to evaluate that output and branch on it, and a way to actually halt execution and mark the run as failed rather than just skip ahead. The Condition step supplies the first piece, checking whether the evaluation result meets whatever bar was defined, such as a metric threshold, and routing execution accordingly, while a Fail step supplies the second, since it's the built-in mechanism for terminating a pipeline run immediately and flagging it as failed, which is what makes it possible to trigger downstream alerting such as an SNS notification tied to that failure state. Together they form the native way to make a pipeline's control flow depend on a step's actual output rather than relying on the pipeline simply running to completion regardless of what the evaluation found. This is more direct than trying to build failure detection into the evaluation step's own code or handling it entirely outside the pipeline, since Condition and Fail steps are the mechanisms SageMaker Pipelines exposes specifically for this. Whenever a scenario wants a pipeline to stop and signal failure based on the result of a prior step, expect the answer to combine a Condition step with a Fail step.
⚠ Common exam trap
Many candidates confuse external monitoring (CloudWatch) or post-step actions (Lambda) with native pipeline control flow, missing that SageMaker Pipelines has a dedicated Condition step for conditional branching and halting execution.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Use a Condition step to check the evaluation result and route to a Fail step if the result indicates failure
SageMaker Pipelines provides a built-in Condition step that evaluates a boolean expression (e.g., checking if evaluation metrics meet a threshold) and then routes execution to different steps. If the condition fails, you can direct the pipeline to a Fail step, which immediately stops the pipeline and marks it as failed. This is the native, event-driven way to halt a pipeline based on step output without relying on external services.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Configure an Amazon CloudWatch Events rule to monitor the pipeline execution status and stop it if the evaluation step fails
Why it's wrong here
CloudWatch Events observes pipeline state after the fact and cannot stop an in-progress execution; stopping requires a ConditionStep inside the pipeline definition. Event rules suit external monitoring, audit trails and alerting on completed executions, not halting a running pipeline at the evaluation step.
- ✗
Register the model in the Model Registry only if evaluation passes, and configure the pipeline to stop if registration fails
Why it's wrong here
ConditionStep evaluates a boolean condition and halts or branches the pipeline; Model Registry registration is a separate governance action, so gating registration does not stop the pipeline on evaluation failure. Registration gating suits promotion workflows where only approved models reach production, not failure-triggered stopping with alerts.
- ✗
Add a Lambda step after the evaluation step that checks the evaluation metrics and sends an SNS notification if the metrics are below a threshold
Why it's wrong here
A Lambda step runs regardless of evaluation outcome, so it cannot itself halt the pipeline; stopping requires a ConditionStep whose condition fails. Lambda-plus-SNS suits custom metric thresholds and notification logic layered on top of an already-branching pipeline, not the mechanism that stops execution.
- ✓
Use a Condition step to check the evaluation result and route to a Fail step if the result indicates failure
Why this is correct
A Condition step evaluates the evaluation step's output against a threshold and, on failure, routes execution to a Fail step that terminates the pipeline and triggers the alert. This satisfies the stem's requirement to stop and notify operations.
Go deeper
Related to this question
About these practice questions
One of 665 original MLA-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
Same concept, more angles
1 more way this is tested on MLA-C01
These questions test the same concept from different angles. Work through them to make sure you can recognise it however the exam phrases it.
Variation 1. A company uses SageMaker Pipelines to automate model retraining. The pipeline runs daily but sometimes fails due to data quality issues. What is the best design to handle this?
medium- ✓ A.Add a data quality check step with Conditional to skip training if data fails.
- B.Use SageMaker Debugger to monitor training.
- C.Use SageMaker Model Registry to track model versions.
- D.Increase the instance size for the training step.
Why A: SageMaker Pipelines supports a data quality check step that can be integrated with a ConditionStep. If the data quality check fails, the ConditionStep can skip the training step entirely, preventing the pipeline from failing due to bad data. This design ensures the pipeline completes successfully (or exits gracefully) without wasting compute resources on training with invalid data.
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This MLA-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the MLA-C01 exam.