Data source dependencies in Terraform - Step-by-Step Execution

Process Flow - Data source dependencies

Start Terraform Apply

↓

Evaluate Data Sources

↓

Check Dependencies

↓

If dependency exists

→Wait for dependency

↓

Dependency resolved

↓

Fetch Data Source Values

↓

Use Data in Resources

↓

Complete Apply

Terraform first evaluates data sources, waits for any dependencies to be ready, fetches their values, then uses those values in resource creation.

Execution Sample

Terraform

data "aws_vpc" "main" {
  filter {
    name   = "tag:Name"
    values = ["main-vpc"]
  }
}

resource "aws_subnet" "example" {
  vpc_id     = data.aws_vpc.main.id
  cidr_block = "10.0.1.0/24"
}

This code fetches the main VPC by tag, then creates a subnet inside that VPC using the fetched VPC ID.

Process Table

Step	Action	Dependency Check	Result	Next Step
1	Start terraform apply	N/A	Begin execution	Evaluate data sources
2	Evaluate data source aws_vpc.main	No dependencies	Fetch VPC info with tag 'main-vpc'	Data source fetched
3	Data source aws_vpc.main fetched	N/A	VPC ID available	Use VPC ID in resource aws_subnet.example
4	Create resource aws_subnet.example	Depends on data.aws_vpc.main.id	Subnet created with VPC ID	Apply complete
5	Apply complete	N/A	All resources created successfully	End

💡 All data sources evaluated and dependencies resolved before resource creation.

Status Tracker

Variable	Start	After Step 2	After Step 3	Final
data.aws_vpc.main.id	undefined	fetched VPC ID (e.g., vpc-123abc)	available	available
aws_subnet.example.vpc_id	undefined	undefined	set to data.aws_vpc.main.id	set

Key Moments - 3 Insights

Why does Terraform wait before creating the subnet resource?

Terraform waits because the subnet resource depends on the VPC ID from the data source. It must fetch the VPC ID first (see execution_table step 3) before it can create the subnet.

What happens if the data source fails to fetch the VPC?

If the data source fails, Terraform cannot get the VPC ID, so it cannot create dependent resources like the subnet. The apply will fail early during data source evaluation (see execution_table step 2).

Can a resource depend on multiple data sources?

Yes, a resource can depend on multiple data sources. Terraform will wait for all data sources to be fetched before creating the resource, ensuring all dependencies are resolved.

Visual Quiz - 3 Questions

Look at the execution table, at which step is the VPC ID first available?

A Step 3

B Step 2

C Step 4

D Step 5

💡 Hint

Check the 'Result' column for when 'VPC ID available' appears.

According to the variable tracker, what is the value of aws_subnet.example.vpc_id after step 3?

A fetched VPC ID

B set to data.aws_vpc.main.id

C undefined

D null

💡 Hint

Look at the 'After Step 3' column for aws_subnet.example.vpc_id in variable_tracker.

If the data source aws_vpc.main had a dependency on another data source, what would Terraform do?

A Ignore the dependency and fetch immediately

B Wait for the other data source to be fetched first

C Fail the apply immediately

D Create resources before data sources

💡 Hint

Refer to concept_flow where dependencies cause waiting before fetching.

Concept Snapshot

Terraform evaluates data sources before resources.
Resources can depend on data source outputs.
Terraform waits for data sources to resolve dependencies.
Data source failures stop resource creation.
Use data source attributes as inputs to resources.

Full Transcript

Terraform applies infrastructure by first evaluating data sources. It checks if data sources depend on others and waits accordingly. Once data sources are fetched, their outputs become available. Resources that depend on these outputs use them to create infrastructure. If data sources fail, Terraform stops before creating dependent resources. This ensures resources have the correct data before creation.