I am trying to install Apache Superset with Helm on Kubernetes (GCP). I am updating the 'values.yaml' locally and then while Gcloud Auth'd and connected to K8 (kubectl) running this command:
helm upgrade superset ~/Documents/superset/ --install -f ~/Documents/superset/values.yaml
Then I run port forward to connect to the K8 Superset via my local browser:
kubectl port-forward service/superset :8088 --namespace default
It assigns a random port that I update in my browser and am able to access Superset with no problems. But I have no dependencies, and I cannot add my Database connection (and test) because when I check my Kubernetes logs I get this error:
Could not load database driver: BigQueryEngineSpec
WARNING:superset.views.base:Could not load database driver: BigQueryEngineSpec
Conversely, I was actually able to add my JSON key as well as the proper SQLAlchemy forms in the UI when I installed Superset only on Docker, and connect to BigQuery no problem.
But with this K8 install, I did it thru the 'values.yaml'. So, I tried to add the content in
extraConfigs:
datasources-init.yaml: |
databases:
- allow_file_upload: true
allow_ctas: true
allow_cvas: true
database_name: bigquery
extra: {
"metadata_params": {},
"engine_params": {},
"metadata_cache_timeout": {},
"schemas_allowed_for_csv_upload": [],
"credentials_info": {
"type": "service_account",................
}
}
.... But I don't think it's this above that's the problem, in the yaml. I think it's the dependencies not loading. ? How can I pull down the /app/pythonpath/superset_config.py file from K8 and view it locally using 'kubectl'?
In the Superset Action Log I see this as well
How do I properly load my dependencies? Is it because I am changing the 'values.yaml' locally and trying to "update" via helm command? Do I have to also change another config file? Or log into the node or cluster itself? I am trying to follow the official documentation. https://superset.apache.org/docs/installation/running-on-kubernetes