Next.js app hosted on AWS Lambda and API Gateway.
Your CircleCI job will need the following environment variables set:
APP_NAME: Deployed name of your application.TF_BUCKET: Name of an AWS bucket you're using to host your Terraform state. I suggest usingtfstate.example.net, replacingexample.netwith your domain name.
In addition, you will need a CircleCI Context called aws-deploy containing an AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY. If you plan to use the same bucket for all your Terraform deployments, you can also move TF_BUCKET into the Context.
Run npx gulp init to set up a local secrets file.
Heads up: If you're switching between environments, Terraform will prompt you with "Do you want to copy existing state to the new backend?" It is important to choose
nohere, since otherwise you will end up with two copies of the same deployment.In other words, if you set up as "dev" and then try to switch to "prod", but choose "yes" at the above prompt, when you attempt to deploy to production you'll only overwrite your dev environment instead.
Next, edit the newly-created .siterc file and add API credentials as shown below:
[google]
clientId=somevalue
clientSecret=somevalueRun npx gulp regenerate to refresh and upload your secrets files.
If you get a "Stage Already Exists" error when deploying, you might have to go through the AWS Console and delete the stage. For this reason, it might be best to deploy each environment once manually before allowing CI to take over. There's probably a way to avoid this error, but unfortunately despite a fair bit of research I haven't found it.
You will need to perform setup once for each environment you plan to deploy; use npx gulp init to reset values.