This Action wraps the Serverless Framework CLI to enable common commands. See their documentation for usage and provider selection.

You'll first have to have a Serverless project as outlined in Serverless's Getting Started. Suppose your provider of choice were Amazon AWS. A workflow could look as follows to install dependencies, and deploy to a Lambda function.

Note: We're first using the NPM Action to install project dependencies, then running serverless deploy via the action.

workflow "Deploy via Serverless" {
  on = "push"
  resolves = ["deploy"]
}

action "install" {
  uses = "actions/npm@master"
  args = "install"
}

action "deploy" {
  needs = ["install"]
  uses = "aaronpanch/action-serverless@master"
  args = "deploy"
  secrets = [
    "AWS_ACCESS_KEY_ID",
    "AWS_SECRET_ACCESS_KEY",
  ]
}

Depending on the Serverless provider chosen, you'll need to supply appropriate credentials. The above example illustrates a typical AWS scenario, but Serverless supports other cloud providers.

Typically, with any cloud provider, their particular authentication environment params are Required.

• SERVICE_ROOT - Optional. To specify a particular subdirectory of your project that contains the Serverless service (the directory with the serverless.yml file) you can specify a SERVICE_ROOT. This action will cd into that directory then execute commands. The default root is . (project root).

To navigate and deploy two services (in this example "Users" and "Admins") in different subdirectories:

action "Deploy Users Service" {
  uses = "aaronpanch/action-serverless@master"
  args = "deploy"
  env = {
    SERVICE_ROOT = "users_service"
  }
  secrets = [
    "AWS_ACCESS_KEY_ID",
    "AWS_SECRET_ACCESS_KEY",
  ]
}

action "Deploy Admin Service" {
  uses = "aaronpanch/action-serverless@master"
  args = "deploy"
  env = {
    SERVICE_ROOT = "admin_service"
  }
  secrets = [
    "AWS_ACCESS_KEY_ID",
    "AWS_SECRET_ACCESS_KEY",
  ]
}