Dry run and Verbose/debug modes for Providers and Provisioners

[delitescere]

Particularly with AWS, there are constantly spurious API glitches that occur. Some actions aren't idempotent. Unfortunately there are also problems with the AWS provider refreshing of state.

I'd like to be able to see what API calls were made by the provider to better diagnose what led to the failure. The same goes for provisioners to make sure the paths and commands are as expected after interpolation occurs.

It would also be beneficial for a dry run mode to show what API calls and provisioner actions would be taken.

This allows for much better visibility under the declarative hood when stuff doesn't work as it should.

[ShonM]

I'd love this. Currently, the only thing I'm seeing from a bad provision is:

Error applying plan:

1 error(s) occurred:

* Script exited with non-zero exit status: 2

Terraform does not automatically rollback in the face of errors.
Instead, your Terraform state file has been partially updated with
any resources that successfully completed. Please address the error
above and apply again to incrementally change your infrastructure.

Just seeing the last command executed would be a great help here

[mitchellh]

This will make it into 0.3 most likely. All the core foundation to support this was put in place in the past week.

[delitescere]

Legend!

Noting that most of the problems' root cause originates with AWS... ;-)

[mitchellh]

Done. See commits above for where it was done. Both local and remote exec provisioners mirror their stdout/stderr.