Exam Professional Data Engineer topic 1 question 47 discussion

You want to optimize the data schema + Machine Learning --> Bigquery. So A

Answer: Should be D - Datastore

For a machine learning-based food ordering service that needs to store transactional data, Cloud SQL is the most suitable option. Here's why: Cloud SQL is a fully-managed relational database service that supports transactional workloads, making it ideal for storing user profiles, account information, and order details. It provides strong consistency and supports complex queries, which are essential for managing and retrieving transactional data efficiently. While BigQuery is excellent for large-scale data analysis, it is not optimized for transactional data storage. Cloud Bigtable is designed for high-throughput and low-latency workloads but lacks the transactional capabilities needed for this use case. Cloud Datastore is a NoSQL database that supports transactions but is generally less powerful than a relational database for complex transactional schemas12. So, the best choice for your needs is B. Cloud SQL.

No idea why so many people go to A. But as transactional data, I think B is correct.

The details of the information definitely look suited to noSql to me, so that means C or D. Datastore is designed for this sort of thing - transactional nosql for an App. I took the question to mean "the machine learning app already exists" so the fact bigquery allows ML isn't relevant. It would be a leap to assume that the ML is done in Bigquery (I have a current Google ML pro cert, and this wouldn't say bigquery to me from that cert)

Cloud SQL is a fully-managed relational database service that supports MySQL, PostgreSQL, and SQL Server. It is well-suited for transactional workloads, allowing you to store structured data with relationships between different entities, such as users, orders, and profiles.

he best answer for this scenario is B. Cloud SQL. Here's why: Relational Data: The information you need to store (user profile, account information, order information) is highly structured and relational. Cloud SQL, being a relational database service, is designed to handle this type of data efficiently. Transactional Workloads: Food ordering involves transactional operations (placing orders, updating user preferences, etc.). Cloud SQL is optimized for transactional workloads, ensuring data consistency and integrity. Ease of Use: Cloud SQL is a managed service, meaning Google handles maintenance, updates, and backups, making it easier to manage than some other options. Integration with Machine Learning: Cloud SQL can easily integrate with other Google Cloud Platform products like BigQuery and Vertex AI, which are crucial for machine learning tasks.

BigQuery is a fully managed, serverless data warehouse that enables scalable analysis of large datasets. It is designed to handle large volumes of data and support complex queries, making it suitable for storing transactional data and performing analytics. With BigQuery, you can optimize your data schema and easily scale as your data grows. Additionally, BigQuery integrates well with other Google Cloud Platform services, including machine learning services, enabling you to build advanced analytics and predictive models on your transactional data.

C It says that the database will be used to store the transactions data. BigQuery is not usually characterized as a data storage system. Also a databased is used for storing transactional Data not a Data Wharehouse.

My choice is C It says "DataBase Schema" not DataWharehouse Schema. It didn't mention if the ML is to be done in the DataBase or not, it just states that a database is to be created.

I asked ChatGPT this question. It first answered Datastore. I said the question asks us to optimize the data schema and Datastore has no schema. Then it answered CloudSQL. I said the question asks about Machine learning aspect. Then it answered Bigquery. I said its a food ordering service and must need low latency Then it answered Bigtable. GPT is clearly not a good tool to use for the prep please avoid it. Its flawed currently that is DEC2023!

A. BigQuery -> Most Probably the answer because for analytic it should automatically be big query. But my question is why not others. I've used BigQuery and I know that it allows schema optimization as well as one of it's feature is built in machine learning. More here:https://cloud.google.com/bigquery/docs/introduction B. Cloud SQL -> I am not being about to find the doc for cloud sql where it talks about being able to optimize the schema and if it can be used for machine learning applications. C. Cloud Bigtable -> Possible answer because there's schema and can be used for machine learning application. Doc: https://cloud.google.com/bigtable/docs/overview D. Cloud Datastore -> Wrong because Datastore doesn’t have a schema. Doc: https://cloud.google.com/datastore/docs/concepts/overview

The right answer is D because Datastore is transactional, scalable, and can deliver an output to ML.

Although you could argue with the formulation of the question, I did also read that this is about a transactional database which BigQuery is not. Thus I would go with Cloud SQL.

Initially I also thought it would be D, but then I re-read & it mentions the whole question is about Data Analytics, hence it has to be A... BigQuery

A sahi hai

to store all the transactional data of the product