Exam AWS Certified Data Analytics - Specialty topic 1 question 90 discussion

The question says least operational overhead...so it should be D

Answer is D AWS Glue can communicate with an on-premises data store over VPN or DX connectivity. An AWS Glue crawler uses an S3 or JDBC connection to catalog the data source, and the AWS Glue ETL job uses S3 or JDBC connections as a source or target data store. https://aws.amazon.com/blogs/big-data/how-to-access-and-analyze-on-premises-data-stores-using-aws-glue/

D is wrong, the requirement is "move the data to a data lake". D doesn't store the source data, but only "save the result to Amazon S3".

I would go with B because of the requirement for minimum overhead. D is also correct but needs more work and services involved, we are using DMS for a data migration scenario and datapipeline for a transformation and movement scenario, it makes sense.

AWS Glue can also connect to a variety of on-premises JDBC data stores such as PostgreSQL, MySQL, Oracle, Microsoft SQL Server, and MariaDB. AWS Glue ETL jobs can use Amazon S3, data stores in a VPC, or on-premises JDBC data stores as a source. AWS Glue jobs extract data, transform it, and load the resulting data back to S3, data stores in a VPC, or on-premises JDBC data stores as a target.

pk349 user is stupid. In all questions posts X: I passed the test. Stupid.

Correct answer is D as AWS Glue can be used to extract from the on-premises data store using JDBC connection, transform data and store the data in S3 with the least operational overhead. Correct answer is D as AWS Glue can be used to extract from the on-premises data store using JDBC connection, transform data and store the data in S3 with the least operational overhead. Option A is wrong as using a customized batch upload process would add to the operational overhead. Option B is wrong as creating an intermediate RDS instance with Data Pipeline jobs added to the operational overhead. Option C is wrong as creating a Redshift cluster would add to the operational overhead.

I think B is not correct. If the goal is to migrate this database to RDS, then is B. But we want to move it to S3, AWS Glue can do ETL. It is less overhead

Answer is D

I think most of us missing the point that we should understand what each solution is designed for. Clearly this is a database migration task so in that case I would defiinetely use DMS, coz its optimized for such tasks. Considering that I would say this is not and ETL task so that Glue ETL isn't an option. Moreover Glue ETL uses "serverless spark platform" and therefore will be expensive for sure. Therefore I go with option B.

C is the correct answer. For any form of analytics, RedShift is a preferred choice. Athena use case is for AdHoc query.

Explanation Correct answer is D as AWS Glue can be used to extract from the on-premises data store using JDBC connection, transform data and store the data in S3 with the least operational overhead. Refer AWS documentation - Glue Analyze On-premises Data Store (https://aws.amazon.com/blogs/big-data/how-to-access-and-analyze-on-premises-data-stores-using-aws-glue/) AWS Glue ETL jobs can use Amazon S3, data stores in a VPC, or on-premises JDBC data stores as a source. AWS Glue jobs extract data, transform it, and load the resulting data back to S3, data stores in a VPC, or on-premises JDBC data stores as a target. Option A is wrong as using a customized batch upload process would add to the operational overhead. Option B is wrong as creating an intermediate RDS instance with Data Pipeline jobs added to the operational overhead. Option C is wrong as creating a Redshift cluster would add to the operational overhead.

We want to reduce operational costs and overhead. A - Sounds good but the beginning is not right, we can do better than customized batch. B - That's too much work, we are not interested in keeping the PSQL Engine, so if we can remove that's step from the equation we are good to go. C - We directly get the table from the On-prem DB with Glue and move it directly to S3. That's good, but we are adding spectrum to the mix, which is a more expensive solution. D - Yes, we are moving directly from on-prem to S3, but using athena to query the data. This is the response.

I believe answer is 'B'

It is close, but I am leaning towards "D"

I believe answer is D. AWS Glue can communicate with an on-premises data store over VPN or DX connectivity. An AWS Glue crawler uses an S3 or JDBC connection to catalog the data source, and the AWS Glue ETL job uses S3 or JDBC connections as a source or target data store. https://aws.amazon.com/blogs/big-data/how-to-access-and-analyze-on-premises-data-stores-using-aws-glue/

The question asks to achieve the goals with "least operational overhead". Answer is D....B cannot be the answer as DMS require EC2 instance to be created for the replication task, this is an operational overhead. Same for A, customized batch process will be an operational overhead. C is not an option as you need to maintain a redshift cluster.