Google Professional Machine Learning Engineer Question 427
Select 3Google Cloud PlatformYou are a Machine Learning Engineer at a company that uses Jenkins for CI/CD pipelines. You need to ensure that your ML training jobs, which leverage Google Cloud ML services, are executed efficiently and automatically upon code changes in a Git repository. Which steps should you take to integrate Jenkins with Google Cloud to achieve this?
- A
Configure a Jenkins pipeline to trigger on Git repository changes using webhooks.
- B
Use the Google Cloud SDK in the Jenkins pipeline to submit ML training jobs to AI Platform (Vertex AI).
- C
Host the Jenkins server on a local machine for better control over the CI/CD process.
- D
Use Jenkins plugins like 'Google Kubernetes Engine Plugin' to manage containerized training jobs on GKE.
- E
Integrate Jenkins with BigQuery to automatically store and analyze CI/CD pipeline logs.
Show answer and explanation
Correct answers: A, B, D
Explanation
To efficiently integrate Jenkins with Google Cloud for ML training jobs, you need to set up a pipeline that triggers automatically on Git repository changes, use tools like the Google Cloud SDK to submit jobs to Vertex AI, and leverage plugins such as the 'Google Kubernetes Engine Plugin' for container orchestration. These steps ensure an automated, scalable, and cloud-native CI/CD workflow for machine learning applications.
- A. Correct.
Correct. Configuring a Jenkins pipeline to trigger on Git repository changes ensures that your pipeline is automatically executed when there are updates to the codebase, which is essential for CI/CD workflows.
- B. Correct.
Correct. The Google Cloud SDK allows Jenkins to interact with Google Cloud services, such as submitting ML training jobs to AI Platform (Vertex AI), which is crucial for running and managing ML workflows.
- C. Incorrect.
Incorrect. Hosting Jenkins on a local machine is not recommended as it may lead to scalability and availability challenges. Instead, you should host Jenkins on a cloud or containerized environment for better reliability.
- D. Correct.
Correct. Using Jenkins plugins like 'Google Kubernetes Engine Plugin' enables efficient orchestration of containerized training jobs on GKE, which is a scalable and robust approach for managing ML workflows.
- E. Incorrect.
Incorrect. While integrating Jenkins with BigQuery may be useful for advanced logging and analysis, it is not directly relevant to triggering and running ML training jobs efficiently.