Project Setup

This guide covers setting up a project to use cdflow2. It assumes you have already installed cdflow2.

Create repository with init command

See init command for details.

Git repository

To deploy a project with cdflow2 it needs a git repo. For example use the GitHub CLI to create a repo:

$ # change "myorg" and "myrepo" to something that makes sense for you
$ gh repo create myorg/myrepo
$ cd myrepo

Config Image

The second TODO is the image used to configure your build & deployment. A good choice if you're starting out deploying to AWS is:

config:
  image: mergermarket/cdflow2-config-aws-simple

Available config images include:

mergermarket/cdflow2-config-aws-simple - a config image for a simple setup with a single AWS account.
mergermarket/cdflow2-config-aws-multi - a config image for a setup with multiple teams deploying to multiple AWS accounts, for larger organisations.

If you need to one or more builds for your release - e.g. a docker container or lambda function - then add to the builds section of cdflow.yaml. For example, to do a docker build and upload it to ECR:

builds:
  docker:
    image: mergermarket/cdflow2-build-docker-ecr

Available build images include:

Setup

Now that you've chosen a config image and build images, you can complete your setup interactively with:

$ cdflow2 setup

The options will depend on the chosen config container.

Terraform

Terraform code should be placed in an infra/ folder - by convention in the following files:

`infra/variables.tf`

The following variables should be defined:

variable "release" {
  type        = "map"
  description = "release metadata: version, commit, component, plus any config container supplied data"
}

In addition each build will result in additional map being provided - e.g. if you have have a mergermarket/cdflow2-release-docker-ecr build under a "docker" key:

variable "docker" {
  type        = "map"
  description = "release data from cdflow2-release-docker-ecr: image_id"
}

In addition you can declare additional environment specific variables - see below for how to provide the values.

`infra/main.tf`

This is where you put your terraform code. In addition to the variables you've declared, you can also access the name of the the environment with the terraform.workspace value.

`infra/outputs.tf`

Any outputs you declare here will appear in the build output for your pipeline.

`infra/backend.tf`

This file contains a partial backend configuration and will be created for you automatically when you deploy. There's no need to add any additional configuraiton as this is provided by the config container you are using. This file can safely be ignored or committed - run the following from the project root to ignore it:

$ echo backend.tf >> infra/.gitignore
$ git commit -m 'ignore generated backend.tf file' infra/.gitignore

Config

To provide the environment specific values of any config variables you have defined (i.e. in infra/variables.tf), add a JSON tfvars file in config/ENV.json - for example:

`config/live.json`

{
  "my_config_name": "config value for the live environment"
}

Build & Deploy

Once this one-off setup has been completed, you can build and deploy your software as follows:

$ # choose a version number:
$ version=1
$ 
$ # build the software
$ cdflow2 release $version
$ 
$ # deploy the software to an "aslive" environment:
$ cdflow2 deploy aslive $version
$ 
$ # deploy the software to a "live" environment:
$ cdflow2 deploy live $version

Choosing a Version Number

It is recommended to use the build number from your CI server and the short git commit, separated by a hyphen. This will allow you to identify the build and the commit that produced it. For example if you're using Jenkins:

# for example "101-d37bfc2"
version="$BUILD_NUMBER-$(git rev-parse --short HEAD)"

Project Setup

.css-1r0pe6c{color:inherit;-webkit-text-decoration:none;text-decoration:none;}.css-1r0pe6c:hover{-webkit-text-decoration:underline;text-decoration:underline;}Create repository with init command

Create repository with init command