A media company wants its clickstream history to stay readable by Spark on another cloud if it ever leaves Google Cloud. Which storage design best keeps the data portable?
Show answer and explanation
Answer: D
Open columnar files such as Parquet, kept in Cloud Storage and exposed through BigLake, can be read by many engines, including Spark running on other clouds, without vendor-specific storage. Choosing portable processing frameworks such as Spark or Beam keeps the logic movable too. Native BigQuery storage and proprietary functions tie the data and the logic to one provider. Spanner is a transactional database and is not designed as a portable analytics archive.
Why the other options are wrong
- A. Native BigQuery storage is efficient for queries, but exporting on request still leaves the data tied to one provider's storage until it is converted to an open format.
- B. Proprietary functions that run only on Google Cloud cannot execute on another cloud, so leaving would require rewriting every transformation. That is the lock-in the company wants to avoid.
- C. Spanner is built for low-latency transactions, not for holding large analytical archives. It also would not provide the open file format that makes the history portable.