Databricks Machine Learning Professional Question 99
Single answerYou are tasked with automating the deployment of a machine learning model after it is promoted to the 'Production' stage in the Databricks Model Registry. The deployment process involves triggering a Databricks Job that updates the serving infrastructure with the new model version. How can you achieve this automation using Databricks features?
- A
Create a Model Registry Webhook that listens for 'Transition to Production' events and triggers the Databricks Job.
- B
Manually check the Model Registry for models promoted to 'Production' and start the Databricks Job when a new version appears.
- C
Set up a Databricks Job to periodically poll the Model Registry for models in the 'Production' stage and deploy the latest version.
- D
Implement a Model Registry Webhook that listens for 'Registered Model Created' events and triggers the Databricks Job.
Show answer and explanation
Correct answer: A
Explanation
To automate the model lifecycle in Databricks, you can use Model Registry Webhooks to listen for specific events, such as when a model is transitioned to the 'Production' stage, and trigger a Databricks Job to handle deployment. This eliminates the need for manual intervention or inefficient polling mechanisms.
- A. Correct.
This is correct. Model Registry Webhooks can be configured to listen for specific events, such as 'Transition to Production,' and trigger a Databricks Job to automate deployment.
- B. Incorrect.
This is incorrect. Manually checking the Model Registry and starting the job defeats the purpose of automation and introduces human error.
- C. Incorrect.
This is incorrect. Polling the Model Registry is inefficient and not recommended when Webhooks can provide event-driven automation.
- D. Incorrect.
This is incorrect. 'Registered Model Created' events occur when a new model is added to the registry, not when it is promoted to 'Production.'