Giters

hashicorp / terraform

Terraform enables you to safely and predictably create, change, and improve infrastructure. It is a source-available tool that codifies APIs into declarative configuration files that can be shared amongst team members, treated as code, edited, reviewed, and versioned.

Home Page:https://www.terraform.io/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Request: Custom Outputs from null_resource/local-exec

qivers opened this issue · comments

commented

Hi

I had a quick search but I couldn't find a feature request for this so apologies if it's a duplicate.

I was wondering if it would be possible to create custom outputs, from a null_resource, running a script in a local-exec?

Our current use case is that we are building custom AMIs. We would like to pass the AMI ID to our Terraform aws_instance and aws_launch_configuration resources without manually updating a variable in tfvars.

So the idea is, we'd have a null_resource with a local-exec which would pull in the latest AMI IDs using a script, and then we could assign the AMI ID outputted from that script to a terraform output;

resource "null_resource" "debian-ami" {
  provisioner "local-exec" {
    command = "python ./ami-finder.py -o debian"
  }
}

output "debiabn-ami-id" {
  value = "${null_resource.debian-ami.local-exec-output}"
}

This script would return the latest debian AMI ID which could then be ingested like any other module output;

ami_id = "${module.debiain-ami-finder.debian-ami-id}"

I realise this might be difficult but wondered if it was something worth considering?

Cheers.

commented

Hi @qivers! It's not possible to do this using null-resource, but the development branch of Terraform 0.7 (on master) now has data driven configuration, and a data source for finding AMIs is on the roadmap for that release!

commented

Hi, @jen20, I'm not able to see how v0.7.0 can fix this issue. This is required essentially for me as I'm using a script to get the private_ips of resources created by auto-scaling group. I need to pass this information to another TF module. So, I need to output this. Is there any way, I can acheive this?

commented

Could @jen20 @apparentlymart explain how to use the new "data sources" feature to run a local script and get the output as a terraform variable? TIA!

commented

We don't (yet) have a data source for running a local command, but for the specific use-case given here (finding an existing AMI) we have aws_ami.

commented

@apparentlymart Yup, like I thought. As far as I think, there is no direct way for this except some hack with null_resource trigger

commented

External data sources get run every time - if you just want to get output once, I made a module for this https://github.com/matti/terraform-shell-resource

commented

Update, now the module works also with remote state setups with an insane trick where I store the outputs in null_resource triggers and then ignore lifecycle changes for the triggers (!) https://github.com/matti/terraform-shell-resource/blob/master/main.tf#L58-L67

commented

Just revisited the solution mentioned above - managed to drop the requirement for external data source (ruby) now.

commented

Will this module find it's way into the core someday?

commented

https://github.com/matti/terraform-shell-resource has been updated so that it updates the outputs if command changes and does not run on every apply

commented

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.