Hello. I’m trying to use Prisma1 for production with AWS. My web API server runs on ElasticBeanstalk and my MySQL DB runs on RDS. I’m trying to deploy Prisma1 on AWS Fargate. In this circumstance, I’m worried about autoscaling. Setting autoscaling-related options on Fargate is the only thing I have to do? Or are there some problems to be considered? Such as what if two different Prisma task on Fargate tries to update the same data in RDS? I’m also considering moving to Prisma2. I heared that Prisma2 can be deployed to AWS Lambda. In this case, are there any problems that I should care?

Hey @Seungsu Kim 👋 I personally have deployed the Prisma Server on Fargate and faced no issues related to scaling or a race condition that you have described. So i think you should be fine deploying to Fargate.

I’m also considering moving to Prisma2. I heared that Prisma2 can be deployed to AWS Lambda. In this case, are there any problems that I should care?

Surely, you can go ahead with Prisma 2 as the stable version has already released and there is a migration guide that we have created as well that you could follow. As for Serverless, yes you can deploy to Lambda. I have personally just created an application that works well.

@Ryan Thank you for your kind answer. Is there any document you referred when deploying Prisma Server on Fargate? The official document (https://www.prisma.io/tutorials/deploy-prisma-to-aws-fargate-ct14) uses CloudFormation and it crates its own VPC and subnets… I’m trying to deploy Fargate in my custom VPC.

Then you would have to edit this template and pass your VPC id wherever you find this

VpcId: !Ref 'VPC'

You can also add that in the

Parameters:

section so as to pass it dynamically on the creation of the resources.