Databricks-ML-Pro ML Ops Practice Question
A team notices that a deployed Model Serving endpoint occasionally returns stale predictions for a subset of customers shortly after the nightly feature refresh completes. The offline feature table is confirmed current. Which is the most likely explanation?
⚠ Common exam trap
The trap here is assuming the online feature store stays synchronized with the offline table without an explicit publish step.
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
✓
The online feature store was not republished after the nightly refresh, so lookups return the previous feature values for those customers.
Feature Store keeps an offline table for training and an online table for serving. Updates to the offline table do not propagate automatically; a publish step materializes them into the online store. When the nightly refresh runs without a subsequent publish, the endpoint keeps reading the prior values, producing stale predictions for affected customers until the online store is refreshed.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
The offline feature table's Delta transaction log has not been checkpointed, so readers see an older snapshot.
Why it's wrong here
Delta readers resolve the current snapshot from the transaction log regardless of checkpoint frequency; checkpoints optimize performance, not correctness. The scenario already states the offline table is current, and the endpoint reads from the online store anyway. Checkpointing therefore cannot explain the stale values.
- ✓
The online feature store was not republished after the nightly refresh, so lookups return the previous feature values for those customers.
Why this is correct
The online store is a materialized copy that only reflects the offline table when a publish runs. If the nightly job updates the offline table but no publish follows, the endpoint continues serving the older values. Republishing after the refresh aligns the online store with the current offline data and eliminates the stale predictions.
- ✗
The model version serving traffic is older than the latest registered version, so predictions lag behind the current model.
Why it's wrong here
Serving an older model version would produce consistently different predictions, not staleness limited to a subset of customers right after a refresh. The symptom is tied to feature values, not model weights. Updating the served version would not correct feature values that the online store has not yet refreshed.
- ✗
The endpoint's autoscaling added replicas that cached outdated feature values during scale-out.
Why it's wrong here
Autoscaling changes compute capacity; it does not cache feature values that persist across requests. New replicas query the same online store, so they would return whatever the online store currently holds. The staleness originates in the online store's contents, not in replica behavior during scaling.
About these practice questions
One of 300 original Databricks-ML-Pro 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 →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Databricks exam blueprint
This Databricks-ML-Pro practice question is part of Courseiva's free Databricks 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 Databricks-ML-Pro exam.