taskcluster/taskcluster-diagnostics

Tests against taskcluster production infrastructure for diagnostics

TaskCluster Diagnostics

NOTE The code in this repository no longer runs on a regular basis and has been retired.

The taskcluster diagnotics component runs tests against the production deployment of taskcluster. This is intend to quickly notify us if for some reason the production deployment breaks. This can happen both when we deploy updates, modify configuration or permissions granted in third-party services, for example permissions for AWS IAM users. With peridic tests we can avoid unpleasent surprises, like realizing that a features in our production deployment have been broken for weeks without anybody noticing.

This also serves as quick way for us to run some tests against the production deployment, which is very useful to do every time we deploy updates. Even though we test our code before deployment, no amount of unit-tests and black box tests will catch the case where we deploy a source update, but forget to update deployment configuration necessary adjustments like new AWS credentials, etc.

Local Development

Create a file user-config.yml

defaults:
  taskcluster:
    credentials:
      clientId:     taskcluster_clientId
      accessToken:  taskcluster_accessToken

  pulse:
    username:       pulse_username
    password:       pulse_password

Service Owner

Service Owner: @ckousik