Executing the command line script:

./build_container.sh

Will build and run the container, then you can navigate to http://localhost:3000 with login with id/password both admin.

To only build the container:

docker build -t volcano-dashboard .

The complete running environment requires another service to expose data store in GNS shared drive via http.

Note: If you've made changes to the plugin, specifying --no-cache option to force rebuild all plugins while rebuilding the container.

The default admin user password is admin, which is not preferred. Spcifying -E GF_SECURITY_ADMIN_PASSWORD={mypassword} to use your own password. Also, -E GF_SECURITY_ADMIN_USER={myuser} to use different admin name.

You may use env.list as env file to make start command shorter and easier.

See env.list for Azure AD example and Github example.

Follow the document to configure Grafana to authenticate users.

Note: The default admin user/password still works after OAuth configured. It's just allows normal user to login with OAuth.

1. Add a new plugin under plugins directory.
2. Follow Grafana plugin development documents to develop the plugin.
3. (Development mode) After it's done, add the plugin's Id (the id in plugin.json) to allow_loading_unsigned_plugins config in the file /grafana.ini in this repo.

Then rebuild the container.

1. Make changes in Grafana console.
2. Find the the dashboard's code by navigating to dashboard's settings window, then select "JSON Model". Copy the JSON code.
3. Paste the copied JSON to the related file under /dashboards in this repo.

The plugin sigrun-gnss which charts datasource from GNS internal shared folder. This requires a GNS internal host to mount the shared folder, then run the container built by this repo's api directory, to expose data via HTTP.

+------------+        +----------+         +-----------+             +--------------------+
| GNS shared |  mount | Docker   |  expose | API       |  http req   | Grafana browser    |
| folder     | ------ | host     | --------| container | <<========= | Sigrun-gnss plugin |
+------------+        +----------+         +-----------+             +--------------------+
                   (eg. kaizan.gns.cri.nz)

In /api directory, run

docker build -t sigrun-gnss-api .

The host (localhost) should have a directory contains similar CSV format as Sigrun's data. Then,

docker run -d  -p {port}:{port} -v {host_dir}:{mount_dir}

to start the service. The exported endpoint will be /data.

Then from plugin part, you'll have to change the hostname:port in plugins/sigrun-gnss/src/datasources.ts. Find the const path in the source file:

const path =  "http://kaizen.gns.cri.nz:8000/data/?"; // assuming the rapidAPI container is running and has been built with the docker compose

Change it to your {localhost}:{port} to make the plugin work.

Run ./build.sh to build the container image and tag it with AWS ECR style. Run ./push.sh to push the image to AWS ECR.

Note: You might need AWS credential configured for your command line to run these.

1. Currently plugins built in dev mode only. There're some linting issue stop building to production mode.

2. Currently plugin sigrun-gnss uses hard-coded URL to the datasource host. A way to configure it with environment variable would be ideal. However, can't find a way to support this in Grafana.