AWS Setup with Serverless v1 After Lambda Runtime End-of-Life

129 views Asked by At

I have a few projects on AWS using the following setup:

• Cognito - (user pools)

• S3 - (frontend)

• Cloudfront - (CDN proxies)

• API Gateway - (REST APIs)

• Lambda NodeJS 14 - (backend logic)

• DATA API - (to make DB queries using data-api-client NPM package)

• Aurora MySQL (compatible with MySQL 5.7.2.08.3) - Serverless v1 (database)

These projects are mainly for development or internal purposes. Hence they are rarely used. Thanks to Serverless v1 database, I can set it to 0 ACU when idle, so basically we are only paying around 1-2$ per month.

Now I learned here that NodeJS 14 is going to deprecate on November 2023. I have a few questions regarding that:

  1. I can still use the existing system after Deprecation Phase 2 as is, right?
  2. What if I want to create the same system for a new project after Deprecation Phase 2? Is there going to be any way to create a deprecated Lambda NodeJS version? Maybe using CLI v2?
  3. If I choose to upgrade to NodeJS 18, then I have to use Serverless v2. In that case, Data API is not supported, so I will need to use RDS Proxy, correct?
  4. Since Serverless v2 can only go to 0.5 ACU, I would need to pay around minimum 40$ per database per month, correct? So in comparison to 2$ for Serverless v1, it becomes 240$ for 6 databases. Is there a better way to optimize the costs?
  5. Is switching to DynamoDB the cheaper option? But then I need to change the DB, Lambda function and all queries for it.
0

There are 0 answers