open-nomad/ui
Buck Doyle 6ff6a41df8
Fix flapping status light test (#6852)
I unintentionally introduced a flapping test in #6817. The
draining status of the node will be randomly chosen and
that flag takes precedence over eligibility. This forces
the draining flag to be false rather than random so the
test should no longer flap.

See here for an example failure:
https://circleci.com/gh/hashicorp/nomad/26368
2019-12-13 09:02:02 -06:00
..
app
config
lib/bulma
mirage
public
tests
vendor
.editorconfig
.ember-cli
.eslintignore
.eslintrc.js
.nvmrc
.prettierrc
.template-lintrc.js
.watchmanconfig
DEVELOPMENT_MODE.md
README.md
ember-cli-build.js
package.json
testem.js
yarn.lock

README.md

Nomad UI

The official Nomad UI.

Prerequisites

This is an ember.js project, and you will need the following tools installed on your computer.

Installation

The Nomad UI gets cloned along with the rest of Nomad. To install dependencies, do the following from the root of the Nomad project:

$ cd ui
$ yarn

Running / Development

UI in development mode defaults to using fake generated data, but you can configure it to proxy a live running nomad process by setting USE_MIRAGE environment variable to false. First, make sure nomad is running. The UI, in development mode, runs independently from Nomad, so this could be an official release or a dev branch. Likewise, Nomad can be running in server mode or dev mode. As long as the API is accessible, the UI will work as expected.

You may need to reference the direct path to ember, typically in ./node_modules/.bin/ember.

The fake data in development is generated from a stable seed of 1. To generate different data, you can include a query parameter of ?faker-seed=2 or any other number in the URL. To turn off the seed and get different data with every load, use ?faker=seed=0.

Running / Development with Vagrant

All necessary tools for UI development are installed as part of the Vagrantfile. This is primarily to make it easy to build the UI from source while working on Nomad. Due to the filesystem requirements of Broccoli (which powers Ember CLI), it is strongly discouraged to use Vagrant for developing changes to the UI.

That said, development with Vagrant is still possible, but the ember serve command requires two modifications:

  • --watch polling: This allows the vm to notice file changes made in the host environment.
  • --port 4201: The default port 4200 is not forwarded, since local development is recommended.

This makes the full command for running the UI in development mode in Vagrant:

$ ember serve --watch polling --port 4201

Running Tests

Nomad UI tests can be run independently of Nomad golang tests.

  • ember test (single run, headless browser)
  • ember test --server (watches for changes, runs in a full browser)

You can use --filter <test name> to run a targetted set of tests, e.g. ember test --filter 'allocation detail'.

In the test environment, the fake data is generated with a random seed. If you want stable data, you can set a seed while running the test server by appending &faker-seed=1 (or any other non-zero number) to the URL.

Linting

Linting should happen automatically in your editor and when committing changes, but it can also be invoked manually.

  • npm run lint:hbs
  • npm run lint:js
  • npm run lint:js -- --fix

Building

Typically make release or make dev-ui will be the desired build workflow, but in the event that build artifacts need to be inspected, ember build will output compiled files in ui/dist.

  • ember build (development)
  • ember build --environment production (production)

Releasing

Nomad UI releases are in lockstep with Nomad releases and are integrated into the make release toolchain.

Conventions

  • UI branches should be prefix with f-ui- for feature work and b-ui- for bug fixes. This instructs CI to skip running nomad backend tests.

Troubleshooting

The UI is running, but none of the API requests are working

By default (according to the .ember-cli file), a proxy address of http://localhost:4646 is used. If you are running Nomad at a different address, you will need to override this setting when running ember serve: ember serve --proxy http://newlocation:1111.

Also, ensure that USE_MIRAGE environment variable is set to false, so the UI proxy requests to Nomad process instead of using autogenerated test data.

Nomad is running in Vagrant, but I can't access the API from my host machine

Nomad binds to 127.0.0.1:4646 by default, which is the loopback address. Try running nomad bound to 0.0.0.0: bin/nomad -bind 0.0.0.0.

Ports also need to be forwarded in the Vagrantfile. 4646 is already forwarded, but if a port other than the default is being used, that port needs to be added to the Vagrantfile and vagrant reload needs to be run.