After deploying a new virtual machine using Terraform, you find that the local script didn't run properly. However, Terraform reports the virtual machine was successfully created. How can you force Terraform to replace the virtual machine without impacting the rest of the managed infrastructure?