prisma deploy

Deploys service definition changes. Every time you're making a local change to any file in the service configuration (including the data model), you need to synchronize these changes with the remote service using this command.

Upon the very first deploy of a service, the command will perform initial data seeding if the seed property in prisma.yml is provided. You can prevent this by passing the --no-seed option.

If no cluster property is specifed in prisma.yml, the command will prompt you to interactively select a cluster as a deployment target. After you selected the cluster, it will write it to prisma.yml as the default deployment target for future deploys. To bring up the interactive prompt again, simply remove the cluster property from prisma.yml manually.

If your deploy results in data loss or brings your project into an incosistent state, the CLI will not perform the deploy and print a warning instead. In these cases, you can force the deploy by using the --force flag.

Usage

prisma deploy [flags]

Flags

-d, --dry-run              Perform a dry-run of the deployment
-e, --env-file ENV-FILE    Path to .env file to inject env vars
-f, --force                Accept data loss caused by schema changes
-j, --json                 JSON Output
-w, --watch                Watch for changes
--no-seed                  Disable seed on initial service deploy

Examples

Deploy service.

prisma deploy

Deploy service and interactively select a cluster as deployment target.

prisma deploy --interactive

Deploy service with environment variables specified in .env.prod.

prisma deploy --env-file .env.prod

Deploy service accepting data loss or brining the project into an incosistent state.

prisma deploy --force