PMLE Automating and Orchestrating ML Pipelines Practice Question

A company is using Vertex AI Pipelines for ML workflows. They want to implement best practices for idempotent components and data passing. Which THREE practices should they adopt?

⚠ Common exam trap

A common misconception in Vertex AI Pipelines is that in-memory data passing is acceptable in containerized components, but the correct pattern is to use GCS URIs and artifact references to ensure idempotency and scalability.

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

✓

Pass large datasets between components using GCS URIs instead of in-memory values.

Option A is correct because Vertex AI Pipelines components exchange data through artifacts, and passing large datasets as GCS URIs (e.g., gs://bucket/path) avoids serializing huge in-memory objects into the pipeline's metadata/execution store, keeping runs efficient and reproducible. Option B is correct because hard-coded paths break portability and reproducibility across environments; using pipeline parameters (or input artifacts) to supply URIs lets the same pipeline definition run against dev, test, or prod buckets without code changes. Option E is correct because idempotent components—where identical inputs always yield identical outputs and re-execution has no additional side effects—are a core Vertex AI Pipelines best practice, enabling safe retries and caching. Option C is not appropriate because passing in-memory objects between components requires serialization, can exceed metadata limits, and undermines the artifact-based data-passing model. Option D is not appropriate because global variables introduce hidden state, are not tracked as pipeline parameters or artifacts, and break reproducibility and caching across pipeline runs.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Pass large datasets between components using GCS URIs instead of in-memory values.
Why this is correct
Passing GCS URIs keeps components stateless and idempotent, since each run reads from an immutable location rather than relying on in-memory state that cannot survive retries or cross-component boundaries. This satisfies the stem's data-passing best practice for large datasets.
✓
Avoid hard-coding file paths; use pipeline parameters to pass URIs.
Why this is correct
Parameterising URIs removes hard-coded paths, so the same component can run against different buckets or datasets without edits, and reruns stay reproducible. This satisfies the stem's idempotency and data-passing best practice by making inputs explicit pipeline parameters.
✗
Read data into memory in the first component and pass the in-memory object to subsequent components.
Why it's wrong here
Passing in-memory objects between components breaks idempotency and reproducibility, since components must serialise inputs and outputs through artefacts or parameters, not Python objects. It tempts because in-process caching feels efficient, but Vertex AI Pipelines runs each component as a separate container, so memory cannot cross that boundary.
✗
Use global variables in the pipeline code to store intermediate results.
Why it's wrong here
Global variables live only in the pipeline's orchestration process, so component containers cannot read them, breaking data passing and idempotency. It tempts as a shortcut for sharing state, but Vertex AI Pipelines requires explicit inputs, outputs and artefacts so each component remains reproducible and rerunnable.
✓
Design components to be idempotent so that the same input always produces the same output.
Why this is correct
Idempotence guarantees that re-running a component with identical inputs yields identical outputs, so pipeline retries or restarts after failure cannot corrupt downstream artefacts or duplicate side effects. This directly satisfies the best-practice requirement for deterministic, safely repeatable components in Vertex AI Pipelines.

Go deeper

PMLE Automating and Orchestrating ML Pipelines Practice Question

Why each option matters

Option-by-option breakdown

Related to this question