Photonic

photonic.mp4

Server

To import secrets from a file:

flyctl --app [app_name] secrets import < ./apps/next/.env

To build Docker image locally:

docker build . -o output -f ./apps/next/Dockerfile

Scale memory to 1024 MB to survive prisma migrate deploy:

fly scale memory 1024 -a [app_name]

DB

Creation:

fly pg create
fly pg attach --app [app_name] [database_name]
# Redploying the app with the updated secret is now required.

Proxy connection URL will be in the format postgres://user_name:password@localhost:5432/user_name

Modifying Tables

1. Make a change to the schema file
2. Apply the change to DB:

Experimentation

3. Apply the change pnpm next db:push
4. When finished, revert the schema change and run pnpm next db:migrate

Permanent

4. Generate and apply migration pnpm next db:migrate

Tip

Deployments run prisma migrate deploy.

Terraform

• Create a new AWS user (non-root)
• Create an S3 bucket called photonic-tfstate. Terraform state will be stored there.

The AWS user must have the following permissions:

• AWSLambda_FullAccess
• iam:AttachRolePolicy
• Inline policy:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "iam:CreateRole",
        "iam:ListInstanceProfilesForRole",
        "iam:PutRolePolicy",
        "iam:DeleteRole",
        "iam:AttachRolePolicy"
      ],
      // Replace '123' with the actual ARN of the user
      "Resource": "arn:aws:iam::123:role/*"
    },
    {
      "Effect": "Allow",
      "Action": [
        "logs:CreateLogGroup",
        "logs:ListTagsLogGroup",
        "logs:DeleteLogGroup"
      ],
      "Resource": "arn:aws:logs:eu-central-1:123:*"
    },
    {
      "Effect": "Allow",
      "Action": "s3:ListBucket",
      "Resource": "arn:aws:s3:::photonic-tfstate"
    },
    {
      "Effect": "Allow",
      "Action": ["s3:GetObject", "s3:PutObject", "s3:DeleteObject"],
      "Resource": "arn:aws:s3:::photonic-tfstate/*/state/terraform.tfstate"
    },
    {
      "Effect": "Allow",
      "Action": "s3:ListBucket",
      "Resource": "arn:aws:s3:::mybucket"
    },
    {
      "Effect": "Allow",
      "Action": ["s3:GetObject", "s3:PutObject", "s3:DeleteObject"],
      "Resource": "arn:aws:s3:::photonic-tfstate/key"
    }
  ]
}

• Create a profile named photonic on the current machine with the AWS credentials for the new user:

# Path $HOME/.aws/credentials
[default]
aws_access_key_id=
aws_secret_access_key=
[photonic]
aws_access_key_id=
aws_secret_access_key=

Usage

# Workspace
terraform workspace select staging

# Download providers
terraform init

# Create/update Lambda (staging)
terraform plan --var-file=staging.tfvars
terraform apply --var-file=staging.tfvars

# Get Lambda ARN
terraform output

Notes

• CFN template
  • Several parameters are hardcoded e.g. AWS user ARN, Lambda name
  • The template URL has to be set as a secret for the Next app.
  • Should update S3 access policy to match bucket name
• Private Linear board: https://linear.app/photonic/team/PHO/all