Skip to content

H2O Open Source Kubernetes operator and a command-line tool to ease deployment (and undeployment) of H2O open-source machine learning platform H2O-3 to Kubernetes.

License

Notifications You must be signed in to change notification settings

h2oai/h2o-kubernetes

Repository files navigation

H2O Kubernetes

Tests

Repository with official tools to aid the deployment of H2O Machine Learning platform to Kubernetes. There are two essential tools to be found in this repository:

  1. H2O Operator - for first class H2O Kubernetes support (README),
  2. Command Line Interface - to ease deployment of the operator and/or deploy H2O to clusters without the operator (README).

Binaries available: Download for Mac / Linux / Windows. Or build from source.

operator

The operator is an implementation of Kubernetes operator pattern specifically for H2O. Once deployed to a Kubernetes cluster, a new custom resource named H2O is recognized by Kubernetes, making it easy to create H2O clusters inside Kubernetes cluster using plain kubectl. The CLI is a binary usually running on the client's side, usable to deploy the operator itself into Kubernetes cluster or create H2O clusters in Kubernetes in cases when the operator itself may not be used. There are also Helm charts available as yet another way to deploy H2O into Kubernetes. Using the operator first and then falling back to CLI/Helm is the recommended approach.

For detailed instructions on how to use each tool, please refer to the specific user guides:

Contributing

Contributions are welcome and encouraged. Please refer to the contributing guide. If you've encountered a bug, or there is any feature missing, please create an issue on GitHub.

License

This project is licensed under the Apache License 2.0.