Exam AWS Certified Database - Specialty topic 1 question 135 discussion

Going for B. since SELECT INTO OUTFILE S3 is available on Aurora. Option C. uses mysqldump who does not dump directly to S3

Anser is B. The data should be readily accessible (e.g. via Athena), so mysqldump is not useful

B. Configure an AWS Lambda function that exports archival data from the DB cluster to Amazon S3 using a SELECT INTO OUTFILE S3 statement, then deletes the archival data from the DB cluster. Schedule the Lambda function to run weekly using Amazon EventBridge (Amazon CloudWatch Events).

If the amount of data to be selected is large (more than 25 GB), we recommend that you use multiple SELECT INTO OUTFILE S3 statements to save the data to Amazon S3 Answer: B

B is correct approch. C mysqldump can not dump into S3 https://aws.amazon.com/blogs/database/best-practices-for-exporting-and-importing-data-from-amazon-aurora-mysql-to-amazon-s3/

B because: 1. Lambda function max run time is 15 min https://aws.amazon.com/about-aws/whats-new/2018/10/aws-lambda-supports-functions-that-can-run-up-to-15-minutes/ 2. SELECT INTO OUTFILES3 is there, and 10GB data per week sounds reasonable to finish copying within 15 min https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/AuroraMySQL.Integrating.SaveIntoS3.html AWS DMS can copy to S3 but option D says continuilly export, while we need weekly https://aws.amazon.com/blogs/database/archiving-data-from-relational-databases-to-amazon-glacier-via-aws-dms/

running a continually DMS job would NOT be operationally efficient, when talk about which, serverless options combined with Lambda and EventBridge would be a much better choice; considering the volume of the weekly archival, the duration would not hit Lambda timeout; however, it seems like more development would be needed for C, cause Select Into Outfile S3 directly integrates with S3. So B.

Option A requires more effort and hence can be ruled out. Option B uses same lambda function for data migration and deletion thereafter. It doesn't work as lambda might timeout. Option C uses mysqldump which is ok but not as efficient as DMS. Option D is the correct solution in my view.

For option D, I will consider how DMS export data export to S3 looks like and also how DMS handle "delete" CDC statements. With DMS option, you need additional tasks to filter data and not easy to maintain.

I think its D https://aws.amazon.com/blogs/database/archiving-data-from-relational-databases-to-amazon-glacier-via-aws-dms/

Lambda functions have 15mins max execution time. If the extract and delete takes longer than 15 mins using a Lambda function won't work. This limitation might rule out option B and C. Option D will work but "continually export the archival data" is not a requirement. Thoughts?

B final answer

B is correct. https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/AuroraMySQL.Integrating.SaveIntoS3.html

Option B correct

Answer should be D? https://docs.aws.amazon.com/dms/latest/userguide/CHAP_Target.S3.html https://docs.aws.amazon.com/datapipeline/latest/DeveloperGuide/dp-object-sqlactivity.html