The easiest way to work on this datasource is by using docker compose approach. First, run the connector:
yarn dev
And then use another terminal to build backend and run Grafana in docker:
yarn server
This will start Grafana on localhost:2999
, anf you will have to manually add the Cognite datasource to Grafana once you log in.
If you prefer the server to bootstrap the connection to Cognite Data Fusion automatically, you need to set the following environment variables:
export CLIENT_ID="..." # Application (client) ID
export CLIENT_SECRET="..." # Client secret
export COGNITE_HOST="..." # e.g. api.cognitedata.com
export COGNITE_PROJECT="..." # e.g. publicdata
export TOKEN_URL="", e.g. "https://login.microsoftonline.com/{tenant}/oauth2/v2.0/token"
yarn
followed by yarn build
should work on systems with a shell.
For debugging and development, use yarn dev
, and for testing use yarn test
.
Build backend for Linux (arm64) with:
mage -v build:linuxARM64
or (amd64) with:
mage -v build:linux
If you are running Grafana natively, you need to build backend and frontend parts of the plugin and link the folder to Grafana's plugin directory.
Run this command to create a symbolic link
in data/plugins
that points to this directory:
cd /path/to/grafana/data/plugins
ln -s /path/to/cognite-grafana-datasource cognitedata-datasource