/grafana_api

Primary LanguageShell

Grafana + API

Exposes a simple API from a database. Then exposes the data the database receives through the API in a dashboard.

First run

$ cp env.sample .env
$ docker-compose up

To-do

  • Configure Grafana container, use this, this and this Done
  • Check this ticket on updates to upsert method so we can use PostgREST as an input too Solved here
  • Provision datasource to Grafana Done, can't use environment variables though
  • Provision dashboard to Grafana Done
  • Create script to add users to Postgres that have INSERT permission to the API schema in the postgres database Done
  • Give select permissions to grafana user in api table Done
  • Separate results boolean from api table Done
  • Add test table and verification (being able to post to the API with generated token) in team creation script Doesn't seem useful anymore
  • Compare audit team with submitted team in a plot in grafana Not needed any more, implemented through policies
  • Count submission lines in a plot too Done, created metrics views
  • Add trigger to write team in submission or use default value (wrap current_setting('request.jwt.claim.team', true) in function?) Not needed any more, submissions table was removed
  • insertar campo accuracy en tabla submissions y quitar de metrics No es necesario, el accuracy se calcula en una vista
  • cambiar vista validate a una vista de submissions donde accuracy = NULL No, agregaremos resultados como tabla estatica y pondremos una vista que calcule accuracy
  • cambiar results.datediff a tipo date con nombre date Done
  • cambiar results.quantity a tipo float con 2 decimales (scale 20, 2 decimales) con nombre billing Done
  • add fingerprint?
  • better python and R clients, more robust
  • add team tables, swap team to team_id in token, join team name in all tables that use team column Done
  • Add JWK generation to team creation script
  • Add send credentials by email to team creation script
  • Add SSL with nginx container and Let's Encrypt
  • add policy: count rows per submission = rows in real Done, but current implementation not working