Aurora Serverless AWS strong read consistency?

Question

Is it possible to get strong read consistency (read-after-write) with Aurora Serverless? I'm using the data api client.

Answer 1

There are 2 falvors of Aurora Serverless:

1. v1 (~2019) : https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless.html
2. v2 (2022) : https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless-v2.html

This question was asked when Serverless v1 was accounced.

Aurora Serverless v1 will always give you read-after-write consistency. It is pretty much a single (writer) instance Aurora provisioned cluster with serverless support to automatically scale (up or down) the compute.

https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless.how-it-works.html

However, Aurora Serverless v2 is very similar to regular Aurora. v2 essentially gives you an new instance type called "db.serverless" which scales in-place, instantaneously. You can add N db.serverless instances into a cluster, and each of them would act like a self-scaling capacity (read replica). Hence, you get the same level of consistency guarantees as regular Aurora, where ONLY THE WRITER gives you read-after-write consistency. Reader nodes and reader serverless capacity will continue to be eventually consistent.

Hope this clarifies things!