mirror of https://github.com/mcuadros/ascode synced 2024-11-23 01:11:59 +01:00

Starlark is very expressive, but, if we're being pedantic, it [isn't quite Turing-complete](https://github.com/google/skylark/blob/master/doc/spec.md#functions). This is desirable because you can't prove that an arbitrary program in a Turing-complete language will finish, so running such programs in a config language implies that maybe the build won't ever finish.

This is a proposal for alternative wording of what we really get from Starlark: expresiveness.

(By the way, very interesting project!)

2019-10-31 11:28:43 +01:00

3.2 KiB

Raw Blame History

AsCode - The Real Infrastructure as Code

AsCode is a tool for define infrastructure as code using the Starlark language on top of Terraform. It allows to describe your infrastructure using an expressive language in Terraform without writing a single line of HCL, meanwhile, you have the complete ecosystem of providers

Why?

Terraform is a great tool, with support for almost everything you can imagine, making it the industry leader. Terraform is based on HCL, a JSON-alike declarative language, with a very limited control flow functionalities. IMHO, to really unleash the power of the IaC, a powerful, expressive language should be used, where basic elements like loops or functions are first class citizens.

What is Starlark?

Starlark is a dialect of Python intended for use as a configuration language. A Starlark interpreter is typically embedded within a larger application, and this application may define additional domain-specific functions and data types beyond those provided by the core language. For example, Starlark is embedded within (and was originally developed for) the Bazel build tool, and Bazel's build language is based on Starlark.

Examples

Simple

Creating am Amazon EC2 Instance is as easy as:

aws = provider("aws", "2.13.0")
aws.region = "us-west-2"

aws.resource.instance(instance_type ="t2.micro", ami="ami-2757f631")

Using functions

In this example we create 40 instances, 20 using ubuntu and 20 using ECS.

aws = provider("aws")
aws.region = "us-west-2"

# It creates a new instance for the given name, distro and type.
def new_instance(name, distro, type="t2.micro"):
    instance = aws.resource.instance(name)
    instance.instance_type = type
    instance.ami = get_ami_id(distro)

    return instance

amis = {}
ami_names_owners = {
    "ubuntu": ["ubuntu/images/*/ubuntu-xenial-16.04-amd64-server-*", "099720109477"],
    "ecs": ["*amazon-ecs-optimized", "591542846629"],
}

# We create the AMI data-source for the given distro.
def get_ami_id(distro):
    if distro in amis:
        return amis[distro]

    data = ami_names_owners[distro]

    ami = aws.data.ami(distro)
    ami.most_recent = True
    ami.filter(name="name", values=[data[0]])
    ami.filter(name="virtualization-type", values=["hvm"])
    ami.owners = [data[1]]

    amis[distro] = ami.id
    return ami.id

# Creates 20 instances of each distro.
for i in range(20):
    new_instance("ubuntu_%d" % i, "ubuntu")
    new_instance("ecs_%d" % i, "ecs")

Using the runtime

ascode comes with a built-in runtime with functions to work with yaml, json, http, etc. Take a look to the documentation.

load("encoding/base64", "base64")
load("http", "http")

dec = base64.encode("ascode is amazing")

msg = http.get("https://httpbin.org/base64/%s" % dec)
print(msg.body())

Installation

The recommended way to install ascode it's download the binary from the releases section.

License

GPL-3.0, see LICENSE

3.2 KiB Raw Blame History