Exam Professional Data Engineer topic 1 question 13 discussion

Initially, thinking D is the best answer but when question is re-re-read, A seems to be correct answer for following reasons 1. Is payment TRANSACTION -- DB should able to perform full blown transaction (updating inventory, sales info etc, though not specified) , not just ATOMIC which DataStore provides 2. Its point-of-sale application, not ONLINE STORE where HIGH number of concurrent users ordering stuff. 3. User Base could grow exponentially - again more users does mot mean concurrent users and more processing power. Its only about storage. 4. Do not want to Manage infrastructure scaling. - Cloud SQL can scale in terms of storage. 5. CloudStore is poor selection for OLTP application - Each property is index - so higher latency Not sure, during exam 2 min is enough to think on various point.. I may be wrong or wrong path ... lets brainstrom..

D seems to be the right one. Cloud SQL doesn't automatically scale

For handling payment transactions in a point-of-sale application with potential exponential growth and without the need to manage infrastructure scaling, Cloud Bigtable would be the best choice.

D seems to be the answer. This is what I think based on my analysis below. POS is OLTP system but now a days NOSQL with ACID properties also are used for OLTP, Cloud sql is good for relational database and it would have been an option here but it clearly says that "you do not want to manage infrastructure scaling". In cloud SQL, which is managed service and not server less, you need to manually do vertical scaling(scale up and scale down). Hence I believe CLOUD SQL is not the option here. I also tried creating a datastore using google cloud console and it gives 2 options now that is cloud firestore in native mode and cloud firestore in data store mode. automatic scaling is available in both where there is no manual scaling up or down is required. Also, both firestore in native and datastore provides ACID properties. Also, firestore is now optimized for OLTP. Please see below https://cloud.google.com/solutions/building-scalable-apps-with-cloud-firestore Though the question only talks about datastore, I am just providing additional information. Considering all what I read through, D is the answer.

D is the hero here. Though Cloud SQL has an upper hand when it comes to transactions(OLTP), it does not autoscale its computing capabilities as compared to datastore. Do visit: https://cloud.google.com/datastore/docs/concepts/overview#what_its_good_for

D is correct: Datastore (currently Firestore in native or datastore mode). It is a fully managed and serverless solution that allows for transactions and will autoscale (storage and compute) without the need to manage any infrastructure. A is wrong: Cloud SQL is fully a managed transactional DB, but only the storage grows automatically. As your user base increases, you will need to increase the CPU/memory of the instance, and to do that you must edit the instance manually (and the questions specifically says "you do not want to manage infrastructure scaling") B is wrong: Bigquery is OLAP (for analytics). NoOps, fully managed, autoscales and allows transactions, but it is not designed for this use case. C is wrong: Bigtable is a NoSQL database for massive writes, and to scale (storage and CPU) you must add nodes, so it is completely out of this use case.

B - not an option C - lack of ACID transactions A - lack of resource automatic scalability D - (correct, IMHO) support ACID, suitable for OLPT and scalable enough

B - not an option C - lack of ACID transactions A - lack of resource automatic scalability D - (correct, IMHO) support ACID, suitable for OLPT and scalable enough

Cloud Datastore (now part of Google Cloud Firestore in Datastore mode) is designed for high scalability and ease of management for applications. It is a NoSQL document database built for automatic scaling, high performance, and ease of application development. It's serverless, meaning it handles the scaling, performance, and management automatically, fitting your requirement of not wanting to manage infrastructure scaling. Cloud SQL, while a fully-managed relational database service that makes it easy to set up, manage, and administer your SQL databases, is not as automatically scalable as Datastore. It's better suited for applications that require a traditional relational database.

This actually is A. I was initially in the D camp, and have spent considerable time reading about it now (circa 1 hour). D is explicitly not suitable for payment transactions, as Datastore supports ACID transactions, but only within entity groups, which are small, localized sets of data. This restriction means that transactions are not suitable for scenarios requiring multi-entity consistency across the entire database. The only two products recommended for payments in the Google ecosystem are Cloud Spanner and Cloud SQL. Is Cloud SQL managed? I'd say it wasn't really, due to the need to configure instances, but that is trumped by the fact it is the only choice suitable for a payment transaction system.

Cloud Datastore is a fully managed, NoSQL document database that is highly scalable and designed to automatically handle large increases in traffic without requiring manual intervention. It's well-suited for applications with a rapidly growing user base.

Firestore extension of Datastore can handle acid transactions and allows autoscaling

A. Requires Transactions.

Transactional = SQL. Anything with payment processors should NOT be NoSQL

Transactional DB

The best choice for this scenario is D. Cloud Datastore. Here's why: Scalability: Cloud Datastore is a NoSQL database designed for automatic scaling. It can handle exponential growth without requiring manual intervention for infrastructure management. NoSQL: This type of database is well-suited for transactional data like payment processing, where the schema can be flexible. Fully Managed: Cloud Datastore is a fully managed service, meaning Google takes care of the underlying infrastructure, maintenance, and updates. Here's why the other options aren't the best fit: Cloud SQL: While suitable for relational data, it might not be the most scalable or cost-effective option for handling massive growth in a point-of-sale application. BigQuery: Primarily designed for data warehousing and analytics, not transactional processing. Cloud Bigtable: A high-performance NoSQL database, but it's more suited for large-scale, low-latency applications and might be overkill for a point-of-sale system.

auto-scale is the key word to unfold the correct answer. D is the the correct answer