Hi all I'm starting to really love Pulumi. It's so fast and simple in most cases. I'm just really stuck on one thing: Outputs and Inputs. I tried reading the docs and it makes sense but it breaks down as soon as I apply it to something practical. For example: I'm trying to pass a Pulumi secret config value to an AWS LaunchTemplate's user_data script. As far as I can understand this variable:

gitlab_runner_token = gitlab_config.require_secret("token")

contains an

Input

. Eventually, in my code, this input gets implicitly converted to a string:

t = Template(open(f"{getcwd()}/modules/aws_asg/user_data.tpl.sh").read())
templated_user_data = t.substitute({
    'gitlab_runner_token' : gitlab_runner_token,
})

I would expect the secret value in the template, but instead it printed this:

Calling __str__ on an Output[T] is not supported. To get the value of an Output[T] as an Output[str] consider: 1. o.apply(lambda v: f"prefix{v}suffix") See <https://pulumi.io/help/outputs> for more details.

. Firstly I don't understand why we're now talking about an

Output

instead of an

Input

, but that aside. I've followed the error message and googled this many times and most of the time the solution seems to be to use the

.apply()

function as it says. However I haven't found a working solution using the

.apply()

function. Mostly it just prints the same error whatever I try. For example, I would expect the following to unwrap the output into a string given the error message but it prints the same error:

t = Template(open(f"{getcwd()}/modules/aws_asg/user_data.tpl.sh").read())
templated_user_data = t.substitute({
    'gitlab_runner_token' : gitlab_runner_token.apply(lambda v: f"{v}"),
})

My hunch here is that the apply function which takes an

Output

does not return a string, but an

Output

. https://www.pulumi.com/docs/reference/pkg/python/pulumi/#pulumi.Output.apply I don't understand why this would ever be useful. 😄 I kind of understand why this is the behavior, it's never converted to a string until somewhere deep in the pulumi code to protect it being a secret, but this makes it totally unusable in many cases where you want the secret before Pulumi starts its templating magic. So how does the apply function really work? Can Outputs ever be converted to strings to be used in templating scripts? Am I better off not using Pulumi secrets for this use case or is it possible? Thanks a lot! Ben

@plain-mechanic-14459 I definitely hear you when it comes to outputs. I struggled with them as well. I previously wrote this which you may have come across when googling: https://leebriggs.co.uk/blog/2021/05/09/pulumi-apply

Thanks for the in depth explanation and link!! This definitely helps, thanks a lot!

yes you’d use

pulumi.Output.all

@billowy-army-68599 That link doesn't work for me

@plain-mechanic-14459 yep, it’s now here: http://pulumi-hugo-origin-pr-3003-2248df98.s3-website.us-west-2.amazonaws.com/docs/concepts/inputs-outputs/

Looks good! @billowy-army-68599