Hey, prefect 1 here. I'm close to put etl flows in...
# best-practices
Hey, prefect 1 here. I'm close to put etl flows in production. I'm currently using eks with a k8's agent on a fargate node to deploy flows. This agents is only used for dev but i'm creating a staging and prod env. Should i deploy a new agent for each env in this cluster or can i stick with one agent that will handle deployed flows with label
"dev", "staging", "prod"
it depends on your preference - check this for some guidance https://discourse.prefect.io/tag/environments in 2.0 this is much easier thanks to profiles and workspaces
specifically for AWS, the AWS best practice is to separate environments through different accounts https://docs.aws.amazon.com/whitepapers/latest/organizing-your-aws-environment/organizing-your-aws-environment.pdf
this way, there's no need to hardcode environment names anywhere in your code or infrastructure configuration - it's all handled on an account level this way
seems like i'll have to get aws certified 😅
thanks i'll have a look at this !
haha my cert actually expires soon, I should probably remove my AWS cert title soon unless I recertify 😂
😂 1
I used Terraform on AWS in order to sync my different AWS Env for Prefect and a Data Lake. It should help you to re build env faster (like a minute or two 😛 ) !
👍 1
🙏 1