The goal of this project is to create an easy to configure dashboard. You will configure the dashboard through the backend and the frontend will adapt.
The frontend is made in ELM and the backend in Python3 using
aiohttp.
- Show a list of dashboards, allow to pick one.
- Update the frontend in realtime whenever an update is done to the dashboard list.
- Show a stub dashboard when one is selected.
- Import dashboards from user-supplied directories.
- Useful for debugging, the
make frontend-debugreloads the page on code change. - The backend serves the frontend on the
/route. The backend and frontend thus run on the same port. - The frontend Elm code is embedded in a small HTML file. This allows using CSS sheets.
- CSS stylesheet is generated with elm-css. It also reloads on code change.
- Generic Jenkins build plugin, used to create graphs in a dashboard. Sends builds to a given coroutine.
- Generic Builds frontend graph, shows builds in a list.
- Example dashboard connecting to a jenkins instance on http://localhost:8090 with username=api and password=api.
- Make builds graph show a svg graph and not a list of builds.
- Make it pretty with some styling [in progress].
- Robust import system for dashboards to avoid name clashes.
- Use aiohttp to retrieve jenkins build info.
Run:
make BACKEND_ARGS='--dashboard-dir=example_dashboards' app-run
This will start the backend, listening on http://localhost:8080.
Now create the example_dashboards directory and fill it with .py
files, you'll see the list of dashboards update in realtime in the
frontend app.
In one terminal, run:
make frontend-debug
In another, run:
make BACKEND_ARGS='--dashboard-dir=example_dashboards' backend-debug
Then go to http://localhost:8000/Main.elm.
Note that the webpage (frontend) code reloads automatically while the python (backend) code doesn't yet.
Note also that, for now, the backend-debug is equivalent to app-run
Makefile target.