Terraform apply -replace flag - Step-by-Step Execution

Process Flow - Terraform apply -replace flag

Start terraform apply

↓

Check -replace flag?

No→Normal apply

Yes↓

Mark specified resource for replacement

↓

Destroy old resource

↓

Create new resource

↓

Apply other changes

↓

Finish apply

Terraform apply checks if the -replace flag is used. If yes, it forces replacement of specified resources by destroying and recreating them during apply.

Execution Sample

Terraform

terraform apply -replace=aws_instance.example

This command forces Terraform to replace the aws_instance.example resource during apply.

Process Table

Step	Action	Resource	State Before	State After	Notes
1	Start apply	-	aws_instance.example: existing	aws_instance.example: existing	Begin terraform apply process
2	Check -replace flag	-	-	-	-replace=aws_instance.example detected
3	Mark resource for replacement	aws_instance.example	existing	marked for replacement	Resource flagged to be destroyed and recreated
4	Destroy old resource	aws_instance.example	marked for replacement	destroyed	Old instance is destroyed
5	Create new resource	aws_instance.example	destroyed	created	New instance is created
6	Apply other changes	-	-	-	Apply any other planned changes
7	Finish apply	-	aws_instance.example: created	aws_instance.example: created	Apply completes successfully

💡 All resources applied; replaced resource destroyed and recreated as requested by -replace flag

Status Tracker

Variable	Start	After Step 3	After Step 4	After Step 5	Final
aws_instance.example	existing	marked for replacement	destroyed	created	created

Key Moments - 3 Insights

Why does Terraform destroy the resource before creating a new one when using -replace?

Terraform must destroy the existing resource first to avoid conflicts and ensure a clean replacement, as shown in steps 4 and 5 of the execution_table.

Does -replace affect resources not specified in the flag?

No, only the resource specified with -replace is destroyed and recreated; other resources are applied normally, as seen in step 6.

What happens if the resource does not exist before apply but -replace is used?

Terraform will simply create the resource without destruction since there is no existing resource to replace, similar to skipping step 4.

Visual Quiz - 3 Questions

Look at the execution_table, what is the state of aws_instance.example after step 4?

A existing

B destroyed

C marked for replacement

D created

💡 Hint

Check the 'State After' column for step 4 in the execution_table.

At which step does Terraform detect the -replace flag?

A Step 2

B Step 1

C Step 3

D Step 5

💡 Hint

Look for the action 'Check -replace flag' in the execution_table.

If you remove the -replace flag, how would the execution_table change?

A Resource would be destroyed but not recreated

B Step 3 would mark resource for replacement anyway

C Step 4 and 5 would be skipped; resource remains existing

D All steps remain the same

💡 Hint

Refer to the concept_flow where no -replace flag leads to normal apply without replacement.

Concept Snapshot

terraform apply -replace=resource
Forces replacement of specified resource.
Destroys old resource, then creates new one.
Other resources apply normally.
Use to fix or refresh problematic resources.

Full Transcript

Terraform apply with the -replace flag forces Terraform to destroy and recreate a specified resource during the apply process. The flow starts with detecting the flag, marking the resource for replacement, destroying the old resource, creating a new one, then applying other changes normally. This is useful when you want to refresh or fix a resource without changing the entire infrastructure. Only the specified resource is replaced; others remain unchanged.