Do I somehow need to initialize pulumi before pulumi.run will work?

Not sure what you mean by

pulumi.run

To execute a Pulumi program, you use

pulumi up

where

pulumi

is the Pulumi CLI. So, confirming a few things: • you have the Pulumi CLI installed as per: https://www.pulumi.com/docs/get-started/install/ • You have a pulumi program written in go - something along the lines of the go example here: https://www.pulumi.com/registry/packages/equinix-metal/ • You have your Pulumi project/stack set up with the necessary Equinix credentials as per: https://www.pulumi.com/registry/packages/equinix-metal/installation-configuration/ • You’ve initialized the stack as per https://www.pulumi.com/docs/intro/concepts/stack/ If so, what output do you get from

pulumi up

?

I was running my script directly, as pulumi.run() is part of a larger function that's called by a http request, is that possible with pulumi up?

There is the automation API package that essentially wraps the Pulumi CLI in an SDK: https://www.pulumi.com/docs/guides/automation-api/ So you could use that to manage Pulumi stacks (and the resources within).

I also see that I never actually got

pulumi new

to run, as it kept asking me to run it in an empty directory

Don’t do a

pulumi new

at this point. You would do a

pulumi stack init

as per the link provided above.

Oh, I ran pulumi stack init and it errored but I guess it did actually work?

┌──(gido5731㉿gtux0)-[~/AutoDeployVM/masterapi]
└─$ pulumi stack init masterapi                                                                                                                                                          255 ⨯
Created stack 'masterapi'
error: no Pulumi project found in the current working directory. Move to a directory with a Pulumi project or try creating a project first with `pulumi new`.
                                                                                                                                                                                               
┌──(gido5731㉿gtux0)-[~/AutoDeployVM/masterapi]
└─$ pulumi stack init                                                                                                                                                                    255 ⨯
stack name: (dev) masterapi
error: stack 'masterapi' already exists

So, although you have some code, you do need to add some files so Pulumi knows how to run it. I would recommend going through one of the getting started sequences here: https://www.pulumi.com/docs/get-started/ to understand the overall flow and main concepts.

I didn't originally plan to make this project with Pulumi, but no one seems to have any way to take my project that I already wrote and tell pulumi to make the files it needs around it

Well, if it’s written using Pulumi SDKs (https://www.pulumi.com/registry/) and is in a directory with a

Pulumi.yaml

(project file) and (possibly) a

Pulumi.XXX.yaml

(stack file for stack named XXX) then

pulumi up

should work.

No pulumi commands want to create a yaml file, they all want either an empty folder or to use a template I don't want. Can I make a yaml manually?

Yes. Look at some of the “go” examples in that link I just provided and emulate what you see there.

No problem, I'll try feeding it a pulumi.yaml and see if it cooperates

The Pulumi SaaS is free for individuals. Or you can use a self-managed backend to store state as per: https://www.pulumi.com/docs/intro/concepts/state/

Right now I have it configured to be logged in to a local backend, as I don't really want any cloud nonsense from some randos service. If I can run my own servers I can run my own authentication

Understood. I just wanted to make sure you knew you did have to

pulumi login

to something.

Yes, I definitely am not a fan of the number of steps required to provide an auth token

Good luck!