Getting Started

Creating your properties project:

Step 1: Creating your config.yml

1. Create your top level directory for your property project.

   mkdir example-properties

2. Create a config directory and inside a config.yml file.

cd example-properties/
mkdir config
cd config
touch config.yml

The config.yml will define essential configurations that the generator needs to create the property json files. In the config.yml file we need three keys set to explain our project to the generator.

1. The environments key. This key will define a array of the environments that we are setting properties for. Environments must be unique and have a one to one mapping with amazon aws regions.
2. The accounts key. This key will define a array of amazon accounts properties will be uploaded to and served in.
3. The environment_configs key. This key will define three things.
   • The one to one mapping or aws regions to environments.
   • The account a environment lives in.
   • Interpolations for a given environment. Interpolations will be explained in a separate section.
4. The vpc key can be added under the environment name in the environment configs if you would like to run multiple environments in the same region. If you specify a VPC in one environment then a VPC must be specified in all environments. You can only specify one VPC per environment.

Here is a example config.yml

environments:
  - 'my-test-env1'
  - 'my-test-env2'

accounts:
  - 123456789012
  - 987654321098

environment_configs:
  my-test-env1:
    region: 'us-east-1'
    account: 123456789012
    interpolations:
      region: 'us-east-1'
      cloud: 'test-cloud-1'
      domain: 'my1.com'
  my-test-env2:
    region: 'eu-central-1'
    account: 987654321098
    interpolations:
      region: 'eu-central-1'
      cloud: 'test-cloud-2'
      domain: 'my2.com'

Step 2: Creating your globals

Globals are properties that get mixed in with service definitions. The hierarchy of the global definition sets the ruling for what that property can supersede. The globals supersedence order is as follows. Any Environmental globals will overwrite account globals. The resultant merge of environmental and account globals overwrite definitions in the top level globals.yml. In short Superscedence from least to greatest is globals.yml, account globals, environment globals. To define globals follow the steps below.

Top level globals

1. Create a globals folder

cd example-properties/
mkdir globals

2. If you would like top level globals then create a globals.yml in your globals folder.

cd globals/
touch globals.yml

3. Define your yaml values in the globals.yml

Account globals

1. Create a folder called the account id of your aws account.
2. In the folder created above create an YAML file named after the account id.
3. Define your account level globals in the YAML file you created above.

Environment Globals

1. Create a folder called environments inside your folder named after your account.
2. Inside the environments folder create a yaml file named after the environment you would like to define globals for. Only environments defined in your config are supported. The environment must also be mapped in the config to the account the sharing the same name as the folder the environment global yaml file lives in.
3. In the newly created environment's yaml file you may define your globals.

Hierarchical Merging

Hierarchical merging follows the following equation (account.yml < service::default) < (region.yml < service::region). So the main thing to note that may be counterintuitive is that service defaults do not overwrite a region default.

Encrypted Properties in Environment Globals

If you would like to pass in an encrypted property to all services in a given environment, you can pass in ckrt encrypted values by having an encrypted block in your environment global file.

property_name_1: value
property_name_2: value
encrypted:
  encrypted_property_1:
    $ssm:
      region: region
      encrypted: encrypted_value
  encrypted_property_2:
    $ssm:
      region: region
      encrypted: encrypted_value

Step 3: Creating your service definitions

Service definitions have the highest level of supersedence and will overwrite matching global definitions. To create you service definitions you need to create the services folder and then the services yaml files.

cd example-properties/
mkdir services
cd services
touch my-service-name.yml

Service definitions consist of three parts default, environments, and encrypted. Encrypted definitions overwrite environment definitions which will overwrite default definitions. Here is a example of a service file. The name of your service file MUST be the same as your service.

default:
  database.host: 'my.database.{domain}'
  database.port: 3306

environments:
  my-test-env-1:
    thread.pool.size: 12
  my-test-env-2:
    thread.pool.size: 8

encrypted:
  my-test-env-1:
   my.encrypted.property:
    $ssm:
      region: us-east-1
      encrypted: PRETEND_ENCRYPTED_PROPERTY_CIPHERTEXT

Adding interpolations

An interpolation is a value that will be dynamically substituted during generation with the correct value for the environment being generated. Interpolations are declared in the config for a given environment. Once declared an interpolation can be used in a property definition by referencing it in braces. If we were to reference the domain interpolation from the example config above we would use {domain}.

Note: values that start with an interpolation must be placed in quotes (ex. "{xxx}.xxx.xxx.{xxx}").

Step 4: Generating Your Properties (Using the CLI)

The bin directory contains the generator.rb cli. An example of running the cli is below. The project_path argument specifies the path to the properties repo we are generating a uploading properties from. You must be able to create a session with s3 to upload.

./generator.rb generate --project_path "~/projects/project-properties" --upload true --upload_account "123456789012" --upload_region "us-east-1" --upload_bucket "propertiesbucket.my-cloud.com"