Professional Cloud ArchitectDemo

Question

You are designing the storage architecture for a financial analytics platform. The platform ingests and retains terabytes of transactional data each day, which supports both real-time fraud detection and long-term historical analysis.

Transaction data from the most recent 30 days must be available with very low latency to the fraud-detection engine. Data more than 30 days old is accessed infrequently for quarterly reports, for which retrieval times of a few seconds are acceptable. All data must be retained for five years to satisfy compliance regulations.

You need to design the most cost-effective solution possible. What should you do?

A Store all transaction data in a Cloud Storage bucket using the Standard storage class for the entire five-year retention period.
B Ingest all data into BigQuery using time-partitioned tables, and rely on BigQuery’s automatic long-term storage pricing for data older than 90 days.
C Configure a Cloud Storage bucket with an Object Lifecycle Management policy to transition data from the Standard class to the Archive class after 30 days.
D Configure a Cloud Storage bucket with an Object Lifecycle Management policy to transition data from the Standard class to the Coldline class after 30 days.

Explanation

Coldline storage is designed for data read or modified at most once per quarter and provides low-latency online access at a lower storage cost than Standard storage. Keeping the most recent 30 days in Standard storage satisfies the fraud-detection latency requirement, while an Object Lifecycle Management policy can automatically transition older objects to Coldline for the five-year retention period. Archive storage is better suited to data accessed less than once per year and has higher access costs.

Learn more

Professional Cloud ArchitectDemo

AstroTutor

Notes

Readiness

QuestionQ287

Community Discussion