cloud-runner

Cloud Runner is a simple microservice that builds and runs a gcloud run deploy command against a given GCP project ID. For flag support please visit the wiki.

Development

Environment Variables

Name	Description	Required	Notes
API_KEY	Validated for Create/Delete operations	✔️
SQL_HOST	SQL host		If not set will default to local sqlite DB
SQL_NAME	SQL database name		If not set will default to local sqlite DB
SQL_PASS	SQL password		If not set will default to local sqlite DB
SQL_USER	SQL username		If not set will default to local sqlite DB

Build

make build

Test

make test

Run Locally

The following will show you how to run Cloud Runner locally and onboard your first account! Running locally will by default create a SQLite DB named cloud-runner.db in the current directory.

1. Run cloud-runner

$ export API_KEY=test
$ make run

2. Create an account. Cloud Runner connects to Spinnaker's fiat (at http://spin-fiat.spinnaker:7003) when deploying to GCP to verify the current user has read and write access to the account. The user is defined in the X-Spinnaker-User request header. When onboarding an account into Cloud Runner make sure to define the read and write groups correctly! If no account field is provided one will be generated in the format cr-<GCP_PROJECT_ID>.

$ curl -H "API-Key: test" localhost:80/v1/credentials -d '{
  "account": "test-account-name",
  "projectID": "test-project-id",
  "readGroups": [
    "test-group"
  ],
  "writeGroups": [
    "test-group"
  ]
}' | jq

You should see the response

{
  "account": "test-account-name",
  "projectID": "test-project-id",
  "readGroups": [
    "test-group"
  ],
  "writeGroups": [
    "test-group"
  ]
}

3. List credentials

$ curl localhost:80/v1/credentials | jq

You should see the response

{
  "credentials": [
    {
      "account": "test-account-name",
      "projectID": "test-project-id",
      "readGroups": [
        "test-group"
      ],
      "writeGroups": [
        "test-group"
      ]
    }
  ]
}

To generate CURL commands to create and monitor a deployment, reference the swagger YAML at api/swagger.yaml.