The Habitat provisioner has been moved into core Terraform (https://github.com/hashicorp/terraform/tree/master/builtin/provisioners/habitat)

terraform-provisioner-habitat

A Habitat provisioner for Terraform

NOTE: This is in a super early state.

• To this point, I have only tested using AWS.
• No validation of the provisioner data is being done yet.
• The only Operating Sytems that have been tested are Amazon Linux 2017.03.0, and Ubuntu 14.04 and 16.04
• Go tests for the provisioner itself are still to come

That being said, please try it out and report any issues you come across!

Installation

Download binaries

Builds for macOS, Linux, and Windows are attached to GitHub releases.

Build from source

git clone git@github.com:nsdavidson/terraform-provisioner-habitat.git
cd terraform-provisioner-habitat
go build

After getting or building a copy of the provisioner plugin, add the following to your ~/.terraformrc (create the file if it doesn't exist)

provisioners {
  habitat = "<path to the plugin binary>"
}

Requirements

• This provisioner will currently only work on Linux targets. As the Habitat supervisor becomes available on more systems, support for those will be added.
• Currently, we assume several userspace utilities on the target system (curl, wget, setsid, tee, etc).
• You must have SSH access as root or a user that can passwordless sudo.

Usage

Example to spin up a 3 node redis cluster:

resource "aws_instance" "redis" {
  ami = "ami-12345"
  instance_type = "t2.micro"
  key_name = "foo"
  count = 3

  provisioner "habitat" {
    peer = "${aws_instance.redis.0.private_ip}"
    use_sudo = true
    
    service {
      name = "core/redis"
      topology = "leader"
      user_toml = "${file("conf/redis.toml")}"
    }
  }
}

Example with service binding:

resource "aws_instance" "web" {
  ami = "ami-123456"
  instance_type = "t2.micro"
  key_name = "foo"
  count = 2

  provisioner "habitat" {
    peer = "${aws_instance.web.0.private_ip}"
    use_sudo = true

    service {
      name = "core/nginx"
      user_toml = "${file("conf/nginx.toml")}"
    }
  }
}

resource "aws_instance" "lb" {
  ami = "ami-123456"
  instance_type = "t2.micro"
  key_name = "foo"

  provisioner "habitat" {
    peer = "${aws_instance.web.0.private_ip}"
    use_sudo = true
    service_type = "systemd"
    service {
      name = "core/haproxy"
      binds = [
        "backend:nginx.default"
      ]
      user_toml = "${file("conf/haproxy.toml")}"
    }
  }
}

Arguments

There are 2 configuration levels, supervisor and service. Values placed directly within the provisioner block are supervisor configs, and values placed inside a service block are service configs. Services can also take a bind block to configure runtime bindings.

Supervisor

• version: The version of Habitat to install. Optional (Defaults to latest)
• permanent_peer: Whether this supervisor should be marked as a permanent peer. Optional (Defaults to false)
• listen_gossip: IP and port to listen for gossip traffic. Optional (Defaults to "0.0.0.0:9638")
• listen_http: IP and port for the HTTP API service. Optional (Defaults to "0.0.0.0:9631")
• peer: IP or FQDN of a supervisor instance to peer with. Optional (Defaults to none)
• ring_key: Key for encrypting the supervisor ring traffic. Optional (Defaults to none)
• skip_install: Skips the installation Habitat, if it's being installed another way. Optional (Defaults to no)
• use_sudo: Use sudo to execute commands on the target system. Optional (Defaults to false)
• service_type: Sets the type of hab-sup service you want to run. Current options are unmanaged and systemd. Defaults to unmanaged.
  • unmanaged: Uses setsid to kick off the habitat supervisor. Not dependent on any init system.
  • systemd: Creates a systemd unit and starts the habitat supervisor service.

Service

• name: A package identifier of the Habitat package to start (eg core/nginx, core/nginx/1.11.10 or core/nginx/1.11.10/20170215233218). Required.
• strategy: Update strategy to use. Possible values "at-once", "rolling" or "none". Optional (Defaults to "none")
• topology: Topology to start service in. Possible values "standalone" or "leader". Optional (Defaults to "standalone")
• channel: Channel in a remote depot to watch for package updates. Optional
• group: Service group to join. Optional (Defaults to "default")
• url: URL of the remote Depot to watch. Optional (Defaults to the public depot)
• binds: Array of binding statements (eg "backend:nginx.default"). Optional
• user_toml: TOML formatted user configuration for the service. Easiest to source from a file (eg user_toml = "${file("conf/redis.toml")}"). Optional

Bind

• alias: The alias for the binding.
• service: The target service to bind.
• group: The target group to bind.

**This format for declaring bindings is optional. It can be used in place of or along side the binds = ["alias:service.group"] method of declaring binds. This format might be easier to manage when populating one or more of the bind parameters dynamically.

Example:

service {
  name = "core/haproxy"
  group = "${var.environment}"

  bind {
    alias = "backend"
    service = "nginx"
    group = "${var.environment}"
  }
}

This block will generate the option --bind backend:nginx.default when starting the haproxy service.