Google Professional Machine Learning Engineer Question 424
Select 2Google Cloud PlatformYou are a Machine Learning Engineer working on a Google Cloud-based project. Your team uses Jenkins to automate the CI/CD pipeline for deploying machine learning models. Recently, the build times for your TensorFlow models have increased significantly, causing delays in deployment. Which strategies should you implement to optimize the build process in Jenkins while leveraging Google Cloud services effectively?
- A
Use Google Cloud Build to offload the build process and integrate it with Jenkins.
- B
Enable distributed training using AI Platform to reduce the model training time.
- C
Cache dependencies and intermediate build artifacts in Jenkins to avoid redundant work.
- D
Migrate the Jenkins master node to a Compute Engine VM with a higher machine type.
- E
Replace Jenkins with Vertex AI Pipelines to completely eliminate build delays.
Show answer and explanation
Correct answers: A, C
Explanation
To optimize Jenkins build times, leveraging Google Cloud Build allows for scalable and efficient offloading of build tasks, while caching dependencies and artifacts in Jenkins can prevent redundant work, further reducing build times. These solutions work within the existing CI/CD framework and address the root cause of the delays effectively.
- A. Correct.
Using Google Cloud Build allows you to offload the build process from Jenkins, leveraging Google Cloud's scalable infrastructure. This reduces the load on Jenkins and improves build times.
- B. Incorrect.
While enabling distributed training using AI Platform can improve model training time, it is unrelated to optimizing the build process in Jenkins.
- C. Correct.
Caching dependencies and intermediate build artifacts can significantly reduce build times by avoiding repetitive tasks, making this an effective optimization strategy.
- D. Incorrect.
Upgrading the Jenkins master node may improve overall performance but does not specifically address the inefficiencies in the build process itself.
- E. Incorrect.
Replacing Jenkins with Vertex AI Pipelines is not a practical solution for optimizing the current build process. Vertex AI Pipelines is designed for orchestrating end-to-end ML workflows rather than replacing traditional CI/CD tools like Jenkins.