Hands-on 2: Snowflake objects

This exercise builds on Hands-on 1: Snowflake provider. Use the same Terraform project and the same SNOWFLAKE_PROFILE environment variable.

Goal

Create the first Snowflake objects with Terraform:

• Database TF_WORKSHOP_<USER_SUFFIX>
• Schema TF_WORKSHOP_<USER_SUFFIX>.LAB

Hands-on 1 established the provider and displayed the current role. This exercise adds the first managed resources.

Step 1: Create terraform.tfvars

Each participant must choose a short, unique suffix based on their username. This prevents participants from creating objects with the same Snowflake names.

Create terraform.tfvars in the project folder:

user_suffix = "<your-username>"

Use only letters, numbers, and underscores. Do not commit terraform.tfvars if it contains personal or environment-specific values.

Step 2: Add the variable

Create variables.tf:

variable "user_suffix" {
  description = "Short user-specific suffix used to make Snowflake object names unique."
  type        = string

  validation {
    condition     = can(regex("^[A-Za-z0-9_]+$", var.user_suffix))
    error_message = "user_suffix may contain only letters, numbers, and underscores."
  }
}

Step 3: Add the resources

Keep the provider, preview feature, current-role data source, and output from Hands-on 1. Add the following resources to main.tf:

resource "snowflake_database" "workshop" {
  name    = "TF_WORKSHOP_${upper(var.user_suffix)}"
  comment = "Database created during the Terraform workshop"
}

resource "snowflake_schema" "lab" {
  database = snowflake_database.workshop.name
  name     = "LAB"
  comment  = "Schema created during the Terraform workshop"
}

The reference snowflake_database.workshop.name creates an implicit dependency. Terraform creates the schema only after the database exists. The unique database separates each participant’s resources, so the schema can use the shared name LAB.

Step 4: Review the plan

terraform fmt
terraform validate
terraform plan

Expect two resources to be created. Review the plan before continuing. plan does not change Snowflake.

Step 5: Apply the configuration

terraform apply

Review the displayed plan and confirm with yes. Verify the result with the Snowflake CLI or Snowsight:

SHOW DATABASES LIKE 'TF_WORKSHOP_<USER_SUFFIX>';
SHOW SCHEMAS IN DATABASE TF_WORKSHOP_<USER_SUFFIX>;

Run the plan again:

terraform plan

The plan should report that no changes are required.

Common pitfalls

The database or schema already exists

Terraform does not initially know about manually created objects. Check that terraform.tfvars contains your unique suffix. Do not run terraform import without first discussing its purpose.

insufficient privileges

The provider login works, but the current role cannot create databases or schemas. Check the role from Hands-on 1 with terraform output current_role and verify that it has the privileges required for this exercise.

The schema is created in the wrong database

Check that the schema uses snowflake_database.workshop.name, not a duplicated or misspelled database name.

Checkpoint

• Keep Hands-on 1 as the known-good authentication baseline.
• Confirm that terraform.tfvars contains your unique suffix before apply to prevent name collisions.
• Predict the resource order before running apply.
• Use a short username or initials as the database suffix; the configuration uppercases it for the database name. Resources inside the database can use shared names.