PMLE Monitoring ML Solutions Practice Question

Your team is using Vertex AI Pipelines to train a model weekly. You want to monitor the pipeline for failures and receive a notification when a pipeline run fails. You have configured the pipeline to send logs to Cloud Logging. What should you do to receive an alert on pipeline failure?

⚠ Common exam trap

The trap here is assuming there is a built-in pipeline failure metric in Cloud Monitoring, while in reality you must derive it from logs.

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

✓

Create a log-based metric that counts error log entries and an alerting policy on that metric.

To alert on pipeline failures, you can create a log-based metric that filters for error logs from Vertex AI Pipelines and then create a Cloud Monitoring alerting policy on that metric. This leverages existing logs and integrates with Cloud Monitoring's alerting system without custom code.

Answer analysis

Option-by-option breakdown

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

✗
Set up a Cloud Function that triggers on pipeline completion and sends an email if the status is failed.
Why it's wrong here
A Cloud Function triggered on pipeline completion could work, but it requires additional infrastructure and code to check the status and send notifications. The native Cloud Monitoring approach with log-based metrics and alerting policies is more integrated and requires less custom code. The scenario implies using existing logging, so the log-based metric is preferable.
✗
Use Cloud Monitoring's built-in Vertex AI Pipeline failure metric to create an alert.
Why it's wrong here
Cloud Monitoring does not provide a built-in metric specifically for Vertex AI Pipeline failures. While some Vertex AI resources have metrics, pipeline failures are typically tracked via logs. Relying on a non-existent metric would not work; you must create a log-based metric from the pipeline's error logs.
✗
Configure the pipeline to publish a custom metric to Cloud Monitoring on failure and create an alert on that metric.
Why it's wrong here
Publishing a custom metric from the pipeline is possible but requires modifying the pipeline code to emit the metric, which adds complexity. Since Vertex AI Pipelines already logs failures to Cloud Logging, using log-based metrics is simpler and more direct. The question asks for the best action given existing logs.
✓
Create a log-based metric that counts error log entries and an alerting policy on that metric.
Why this is correct
Vertex AI Pipelines logs pipeline execution events to Cloud Logging. By creating a log-based metric that filters for error-level logs from the pipeline, you can then create an alerting policy that triggers when the metric exceeds a threshold (e.g., >0 errors). This is the standard way to alert on pipeline failures using logs.

Go deeper

PMLE Monitoring ML Solutions Practice Question

Why each option matters

Option-by-option breakdown

Related to this question