Skip to main content

Datalayer 0.0.1 on Kubernetes

· 2 min read
Eric Charles

Connecting a Notebook to Azure Directory is a piece of work. Building a complete scalable Data Science Platform on Kubernetes is another piece which is more challenging to achieve. The Datalayer Science Platform is designed to be Simple, Collaborative and Multi Cloud and is particulary suited for Data Science exploration teams.

Datalayer Platform version 0.0.1 is released and supported on AWS Amazon Cloud. Today, you can create Kubernetes cluster on the AWS Amazon cloud and deploy as many as you want Apache Zeppelin based Notebooks with Apache Spark (data analytics) and Apache Hadoop (distributed file system) running natively on Kubernetes. You will also get a Kubernetes Dashboard with Heapster to monitor resources usage. We also ship the premises of Kuber Plane, a user interface to control your cluster which also acts as a collaborative Data Science Notebook to share datasets analysis with authentication via Microsoft Azure Directory and Google OAuth.

We are looking forward to your feedbacks on the current version and inputs on our release plan to fit your needs, especially in terms of Applications, Storage, Cloud... you want to use.

You are welcome to contribute on this project as a User or as a Developer on the base source code available in our Github repositories https://github.com/datalayer.

Control Plane

Cluster Scaling

Cluster Reservations

Note

Notebook Cover

Scratchpad