Hello everyone, I wanted to ask a question around automating Point In Time Recovery utilizing AWS RDS. Right now I am deciding to automate our incident response utilizing PITR through CircleCI.I am thinking of doing this with either Terraform or simply writing a bash/node script that will utilize the AWS SDK/CLI. Is there a strong reason for me to utilize Terraform over simply writing a script that will handle all the configurations/automation process? We have a lot of our infrastructure created and managed through terraform so I can see potential issues with malformed configurations if I do not use terraform.

IMO TF isn't a great tool for database administration. We looked at this and we're going with a script-based solution. I don't know your use case but I'd be more concerned about not having a human in the loop for a PITR.

Restoring is a simpler activity and can be done with terraform, the problem I see with custom scripts is maintainability and tribal knowledge. If the team is really good at documenting, script is probably a better choice. But you don't want to end up with a bunch of scripts that future team members are not able to figure out. That being said, maintaining terraform states is also a mess that will come up after you have a bunch of automation in place. Just make sure you account for that, and either of these will work fine.

Thanks @Mark Cheshier and @Vivek Dwivedi. I ended up going the route with AWS SDK/node route and I think it will be a much better choice as all of our developers know node as compared to only a handful of devs understanding Terraform.