pecastel
Cisco Employee
Cisco Employee

Overview

Red PNDA is aimed to be a smaller, simpler subset of PNDA; it provides a set of components providing a PNDA-like environment for development, education and demonstration. It’s more lightweight and designed to run on a laptop, enabling users to get familiar with the core data-ingest mechanism of PNDA (Kakfa/AVRO), as well as the data-exploration tools Jupyter, OpenTSDB and Grafana. It includes Apache Spark and Hbase but doesn’t include the ‘heavy’ components such as the Hadoop infrastructure and distributed processing.

Scenarios

  • Scenario 1: Start PNDA
  • Scenario 2: Explore Jupyter

Components

•       Console Frontend - https://github.com/pndaproject/platform-console-frontend

•       Console Backend

•       Platform Testing

•       Platform Libraries

•       Kafka 0.11.0.0

•       Jupyter Notebook

•       Apache Spark 1.6.1

•       Apache Hbase 1.2.0

•       OpenTSDB 2.2.0

•       Grafana 4.3.1

•       Kafka Manager 1.3.3.6

•       Example Kafka Clients

•       Jmxproxy 3.2.0

Features

Red PNDA

  • This framework provisions a minimal set of the PNDA (pnda.io) components to enable developers writing apps targeted at the full PNDA stack, to experiment with the PNDA components in a smaller, lightweight environment. Data exploration and app prototyping is supported using Jupyter and Apache Spark.
  • Complete Documentation is available at https://github.com/pndaproject/red-pnda/blob/develop/README.md

Apache Kafka

Jupyter Notebook Data Exploration

Some sample notebooks available in Red PNDA:

  • A network-related dataset (BGP updates from the Internet) and an accompanying tutorial Juypter notebook named ‘Introduction to Big Data Analytics.ipynb’.
  • ‘red-pnda-anom-detect.ipynb’ notebook is an Python implementation of a simple mu-sigma-based detection algorithm to identify packet losses in XR telemetry data captured from a simple multiple host topology. Such packet losses are often indications of unintentional black-holing of traffic. A small accompanying dataset is also provided to help people get started.
Time Series Data Exploration
  • Send data to OpenTSDB (a time series db) and visualize it in Grafana. Can be done via Jupyter notebook. See ‘red-pnda-anom-detect.ipynb’ jupyter notebook as an example.

Resources