lightstep-action-predeploy

The lightstep/lightstep-action-predeploy action provides insight into the health of production service(s) during software delivery alongside on-call and error information.

This Javascript-based action can be used when a pull request is approved by GitHub reviewers and uses APIs to provide a summary of deployment risk ahead of a code change going to a production environment.

📖 Follow the tutorial on setting up this action on Lightstep's Documentation Portal

Requirements

GitHub

• Repository that corresponds to a service running in a production environment
  • Github action workflow for the repository that uses the Lightstep GitHub action
  • API token(s) defined as secrets in the repository configuration settings
  • .lightstep.yml configuration file in root directory of repository

Lightstep

• Instrumented service(s) running in a production environment (create account here)
  • Deployment tracking configured for the service
  • Streams created for instrumented service
  • Alert threshold(s) defined for streams
  • Lightstep API key

Usage

This action can be run on ubuntu-latest GitHub Actions runner as a step in any GitHub Action workflow.

Most users will also want to specify a list of specific conditions to check in the lightstep.yml file (see example below).

Example Workflow

This workflow uses the action to add a comment to a pull request after a review has been submitted.

👉 Try out this workflow on your repository

.github/workflows/predeploy.yml

on:
  pull_request_review:
    types: [submitted]

jobs:
  deploy_check_job:
    runs-on: ubuntu-latest
    name: Verify Pre-Deploy Status

    steps:  
      # Using checkout is required if reading from a `.lightstep.yml` file in the repo
      - name: Checkout
        uses: actions/checkout@v2

      # Run checks and output status as PR comment
      - name: Lightstep Pre-Deploy Check
        id: lightstep-predeploy
        uses: lightstep/lightstep-action-predeploy
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        with:
          lightstep_api_key: ${{ secrets.LIGHTSTEP_API_KEY }}
          pagerduty_api_token: ${{ secrets.PAGERDUTY_API_TOKEN }}
          rollbar_api_token: ${{ secrets.ROLLBAR_API_TOKEN }}

Additional Examples

A fully working example is available in the hipster-shop repository.

Inputs

Inputs can be specified using any combination of a .lightstep.yml file in the root of the repository using the action, as inputs to the action, as environment varables.

API keys are always passed as inputs or environment variables.

The following are required:

Action Input	.lightstep.yml	Env var
lightstep_organization	organization	LIGHTSTEP_ORGANIZATION
lightstep_project	project	LIGHTSTEP_PROJECT
lightstep_api_key	N/A	LIGHTSTEP_API_KEY

Additional inputs (optional):

Action Input	.lightstep.yml	Env var
pagerduty_api_token	N/A	PAGERDUTY_API_TOKEN
rollbar_api_token	N/A	ROLLBAR_API_TOKEN
rollup_conditions	N/A	N/A

rollup_conditions will collapse all conditions to a single row of the output table if set to true.

.lightstep.yml Configuration

This action reads configuration from a special file you add to the root of your repository.

Here's an example .lightstep.yml file that specifies PagerDuty and Rollbar-specific configuration:

Optionally, you can specify a list of condition ids to check status of the project with. The default behavior is to check all conditions on the project.

organization: -13ac9ef7
project: hipster-shop

# optional - specifiy specific conditions to check
conditions:
  - YfDzWxGL
  - gLFxv0B1

# optional integrations
integrations:
  rollbar:
    environment: production
    account: lightstep
    project: hipster-shop
  pagerduty:
    service: PA0B3RR

Outputs

• lightstep_organization - Lightstep organization associated with an account

• lightstep_project - Lightstep project associated with an organization

• lightstep_predeploy_md - Markdown-formatted summary of pre-deploy checks

• lightstep_predeploy_status - Status of pre-deploy checks: ok, warn, error or unknown

Notes

This action does not automatically generate a pull request comment after running.

To add the output of this action to generate a pull request comment, use the lightstep_predeploy_md output with another action like unsplash/comment-on-pr.

License

Apache License 2.0