-
Notifications
You must be signed in to change notification settings - Fork 187
Open
Description
Hi, I've installed FfDL in a completely offline kubernetes cluster:
- Imported all the necessary docker images to each cluster node.
- Inited tiller with specified image so it won't pull from the Internet.
- Installed FfDL using helm.
- Trained the example model according to your instructions.
Everything worked fine, and I've got the training results, but Grafana showed nothing but mostly a 'no data points' hint on its panel.
Four dashboards:




And I can't find any useful Prometheus or Grafana logs.
BTW, I've commented out the env variable 'GF_INSTALL_PLUGINS' in spec of container 'grafana' in templates/monitoring/prometheus-deployment.yml, for it would try to download from the Internet.
Any hint on what's missing? Thanks!
Metadata
Metadata
Assignees
Labels
No labels