Resources vs Data Sources

​

Key differences at a glance

​

Examples (syntactic comparison)

# resource example — Terraform will create and manage this VNet
resource "azurerm_virtual_network" "example_vnet" {
  name                = "example-vnet"
  location            = "eastus"
  resource_group_name = "example-rg"
  address_space       = ["10.0.0.0/16"]
}

# data source example — Terraform reads an existing VNet's attributes without managing it
data "azurerm_virtual_network" "existing_vnet" {
  name                = "existing-vnet"
  resource_group_name = "existing-rg"
}

​

When to use each

• Use resource when:
  • Terraform should create, update, or delete the object.
  • You want the object tracked in Terraform state.
  • You need full lifecycle management and drift detection.
• Use data when:
  • The object is managed outside this Terraform configuration (or by another team).
  • You need to reference attributes (IDs, names, subnet lists, AMI IDs).
  • You want to avoid Terraform creating duplicate or conflicting infrastructure.

​

Practical tips

• Prefer data sources for shared resources (e.g., centrally managed networks, shared subnets).
• Avoid using data sources as a workaround to hide unmanaged drift — if you need to manage something consistently, convert it into a resource.
• Remember that data lookups may introduce dependency ordering; reference them explicitly to ensure correct evaluation.

​

References

• Terraform: Resources
• Terraform: Data Sources
• Azure Provider — azurerm Virtual Network