Question

A DevOps engineer must run a shell script on a newly created EC2 instance after Terraform provisions it. They have decided that user_data cannot be used for this task, so they add a remote-exec provisioner to the aws_instance resource. During apply, Terraform reports that it cannot establish a session to run the commands. What is the MOST likely cause of this failure?

Accepted Answer

The provisioner is missing a connection block specifying how Terraform should reach the instance (e.g., SSH type, host, user, and credentials).

Answer

remote-exec provisioners can only run during terraform destroy, so it will never execute during a create apply.

Answer

A depends_on argument must be added to the provisioner block before Terraform will attempt any connection.

Answer

remote-exec requires a separate null_resource because provisioners cannot be nested inside a managed resource.