Beta 0.3.0: Eventador Notebooks
Integrating Jupyter Notebooks into our platform

  by Kenny Gorman, Founder and CEO | 06 Oct 2016

Real-time data is only as good as your ability to analyze and use that data. We wanted a powerful yet simple interface to Eventador data pipelines. Notebooks have been very popular inside the data science community for some time, and they are a natural fit for Eventador.

Today we are releasing Eventador Beta 0.3.0 which includes Eventador Notebooks.

Eventador Notebooks are an automatically deployed notebook environment to make real-time data analysis, experimentation, and manipulation easy. We believe Eventador Notebooks will unlock new levels of usefulness and value out of Eventador.io, and enable your data to be that much more powerful.

Easy setup

Eventador Notebook setup is automatic. A private Notebook server is dynamically provisioned on AWS with every Eventador Deployment. They are deployed inside the same VPC and protected by the same IP whitelist as the rest your Deployment. It’s all built and ready for multiple users to immediately analyze data.

Powerful streaming data analysis

Data Scientists, Analysts, Reporting Engineers, or anyone needing access to the data can build, perform analysis, experiment, save, notate, and share Notebooks. You can dynamically sample a stream straight from the Kafka endpoint, or, use the PipelineDB stack to access real-time aggregated data via continuous views. Or maybe sample data into Pandas dataframes or mutate the data and feed another stream - It’s all built in.

Built-in helpers

Notebooks are awesome, but connecting to your streaming environment could be a hassle. We made this easy by providing a set of helper functions that automatically understand the connection context of your Eventador Deployment. Fetching from topics, or publishing to topics, or running some SQL against the PipelineDB stack is just a simple helper method. It’s all bundled in and ready to go. Updating the helper functions is a simple one-liner, so helpers can be delivered to users easily.

For example:

e.kafka                 Return a PyKafka connection object
e.kafka_connect_string  Return Kafka connect string to use in applications
e.kfetch(topic)         Return sample documents (default: 5) from topic [ex: e.kfetch('my_messages')]
e.publish(topic, doc)   Publish a single document (from a dictionary) to topic
e.topics                List topics in Kafka

The full list of helpers can be found by typing e.help() inside your Eventador Notebook.

Graphs and Charts

Creating graphs and charts is similar to any other notebook environment. The powerful matplotlib library is automatically loaded. Creating reports, showing real-time graphs, plotting statistics from the fire hose of real time data is really easy and powerful.

graph

Based on the open source Jupyter Project

Eventador Notebooks are based on the open source Jupyter Notebook project. If you aren’t already familiar with all the capabilities on Jupyter, you can check out the documentation here. We are active Jupyter supporters!

Getting Started

To try out Eventador Notebooks, sign up for a Eventador Beta account, sign in, and create a Deployment. Our Getting Started Guide leads you quickly through the process. If you already are using Eventador, we will attach a new notebook server to each of your deployments. Every account has an example notebook called “Brewery Example” with examples.