Exam AWS Certified Database - Specialty topic 1 question 59 discussion

It is a close call between B and D. However, not everyone talks about the actual API call that can be used for Aurora Serverless, which is called as AutoPause. More details in the link. https://docs.aws.amazon.com/AmazonRDS/latest/APIReference/API_ScalingConfiguration.html So it should be D

A - meets the basic requirement. But its not cost effective. B - Good option. Saves cost on the storage layer with copy-on-write feature C - Meets the requirement, bit not cost effective. D - Most cost effective. https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless.how-it-works.html#aurora-serverless.how-it-works.pause-resume

I go for D. B looks promising. However, the requirement does not state the data in the clusters is from the same source. Also, keeping the DBs up only 8 hours a day is great for Serverless with auto-pause.

A is also viable solution.

Definitely D since that is the only option that also allows to save on compute by stopping unneeded resources outside of the 8 hours/5 days a week they are needed

- 'Most Cost effective' is the Key - Using clone db you can save on the storage cost but still you have to pay for Compute resources which are going to be charged more than 8Hrs. - Aurora server less reduces the instance cost which will be the most cost saving

Most-Cost effective solution is the key here. REfer to https://aws.plainenglish.io/aurora-database-clones-what-they-are-and-when-to-use-them-b82be9d60309 Pricing Now let’s talk about the pricing for creating and using Amazon Aurora database clones. We get billed per hour for the database instances provisioned as part of the clone database. The pricing is based on the DB instance class that we select for the clone. We are also charged for any addition storage that we use as we make edits to the clone database. Option D is full copy of the DB and definitely will cost more than Option B.

Answer should be D

Answer:D

Option B is correct. Question talks about destroying the cluster after 2 weeks, With aurora you will not pay for the compute capacity but you will pay for the storage capacity. If Option D had the delete option, I could have happily choose Option D. The advantage of Option B, you are not paying additional Storage as clone use the same storage + delta changes.

It is necessary to support operation for 8 hours.

single dev & test instanace -> Aurora clone -> delete clones at end of dev cycle

Even if u pause serverless cluster storage, doesnt address the destroyed after 2 weeks requirement. I go with B.

Amazon Aurora now allows you to create clones between Aurora Serverless v1 and provisioned Aurora DB clusters to enable quick sharing of data.

I vote for B: There isn't an "option to pause the compute capacity on the cluster and set an appropriate timeout." in the RDS console for Serverless.

I prefer B. Aurora Serverless is not compatible to all Aurora provisioned engine version. However, you can do clone with most engine version. Meanwhile, I also consider the performance while restoring snapshot to Aurora Serverless.

It is D