Google Professional Data Engineer
Get started today
Ultimate access to all questions.
Ultimate access to all questions.
In the us-central1 region, machine learning engineers are working with approximately 200 TB of data intended for training machine learning models. Only a small subset of this data is used for each training session, and the data is accessed about once per month. The goal is to minimize storage costs while ensuring the data remains highly available and reliable. What storage solution would you recommend?
Explanation:
Cloud Storage Nearline storage is the optimal choice for data that is accessed approximately once per month, offering a balance between cost and availability. While Cloud Storage Multi-region provides high availability, it is more expensive and unnecessary for data accessed within a single region. Persistent disks, whether SSD or balanced, are more suited for use with virtual machines and would introduce additional costs and operational complexity compared to Cloud Storage, a managed service designed for such use cases. For more details, refer to Cloud Storage documentation.