OpenAddresses Deploy Tools for Cloudformation
- Store and manage AWS credentials locally for one or more AWS accounts
- Create, Update, & Delete CF based stacks from the terminal
Run the following
npm install -g @openaddresses/deploy
This will make the deploy
command available globally
Before you can make changes to any of the underlying infrastructure you must first authenticate the deploy cli
To do so run:
deploy init
and follow the prompts for your credentials.
Note the profile name
prompted for should ideally match the profile name as set in your AWS credentials
file located at ~/.aws/credentials
If the profile name is found in your AWS credentials file, the
credentials from the file will be linked. If it does not exist, you will be prompted for a set of credentials
Once finished run
deploy
to see a full list of options
Note: The credentials file can be found in the ~/.deployrc.json
file
If an account uses tags for billing, the following can be used in the ~/.deployrc.json
file to ensure that
tags are attached to all stacks deployed to that profile
Key | Notes |
---|---|
region |
Default AWS Account Region |
Key | Notes |
---|---|
tags |
Cloudformation Tags to apply to stack |
github |
Github API token for updating deployment status |
Tags can be added to all resources in a stack that is deployed. The tags array can contain either Keys as strings, or Key/Value objects.
Key as strings will be automatically populated for each resource of the CF template and the Value added as a Parameter of the stack
Key/Value objects will be automatically populated for each resource except that no Parameter will be added - the Value will be used directly
{
"<profile_name>": {
"region": "<region>",
"tags": ["Project", {
"Key": "Owner",
"Value": "ingalls"
}, "Client", "<another tag>"]
}
}
If you run deploy init
for a single AWS profile, all resources created with the tool will automatically
be deployed to this "default" account.
If multiple AWS profiles are created via batch init
, then you will either need to use
the --profile <name>
flag when interacting with the API, or to specify the profile in your .deploy
file
The ./deploy
file is created in the root directory of the git repo and follows the following format:
{
"profile": "name of AWS Account profile",
}
By default, if a Dockerfile
is found in the project root, the ECR will be queried before deploy to ensure
the image has been built. IE: a git repo named my-project
would look for an image called my-project:<Git Sha>
.
If you are building multiple docker images, or want to disable this feature, the following options are available
via the .deploy
file.
Disable ECR Check
{
"artifacts": {
"docker": false
}
}
Custom Pre/Postfix
{
"artifacts": {
"docker": "custom-ecr:{{gitsha}}"
}
}
{
"artifacts": {
"docker": "{{project}}:prefix-{{gitsha}}-postfix"
}
}
Multiple Images
{
"artifacts": {
"docker": [
"{{project}}:backend-{{gitsha}}",
"{{project}}:frontend-{{gitsha}}"
]
}
}
Lambda uploads are not watched for by default. Set artifact listeners via your .deploy
file
using the examples below to ensure that they are present on s3 before deploy.
Disable Lambda Check (Default)
{
"artifacts": {
"lambda": false
}
}
Single Lambda
{
"artifacts": {
"lambda": "<bucket>/{{gitsha}}.zip"
}
}
Multiple Lambdas
{
"artifacts": {
"lambda": [
"<bucket>/deploy-lambda/{{gitsha}}.zip",
"<bucket:>/{{project}}-{{gitsha}}.zip"
]
}
}
Tags can be added to all resources in a stack that is deployed. The tags array can contain either Keys as strings, or Key/Value objects.
The format to adding project specific tags is identical to profile tags
as defined in the ~/.deployrc.json
file. See this documentation
for complete information on tag formatting.
{
"tags": ["Project", {
"Key": "Owner",
"Value": "ingalls"
}, "Client", "<another tag>"]
}