IKATS Sandbox

What is IKATS?

IKATS (Innovative toolKit for the Analysis of Time Series) is the result of a R&D Data Science Project and is carried out by CS Systèmes d'Information in collaboration with the AMA LIG team.

IKATS provides a ready to use toolkit for the handling, exploratory analysis and visualization of large volumes of industrial Time Series (TS) data within a single big data framework.

The project website is hosted at : https://ikats.org

There is also a community and a commercial support.

What is the IKATS Sandbox (and what it's not)

• IKATS was primarily designed to work on cluster of machines and not a single desktop.
• Our IKATS Sandbox repository allow you to use it nevertheless, but user must be aware that its performance is not representative of the tool.
• Especially, some algorithms which use Apache Spark could monopolize all desktop resources and make irreversibly freeze it, depending on the amount of data processed at a time.
• Data persistence: all data imported and generated in ikats are persisted as long as /var/lib/ikats/docker_bindings directory is not manually deleted or altered.

Installation on Linux

Prerequistes

The IKATS Sandbox is known to work on that system :

• Linux version: Ubuntu 16.04
• 8 Go of RAM
• Dual core CPU @ 2.40GHz
• Docker CE 17.12.1-ce
• docker-compose 1.19.0

Install steps for Ubuntu

Tooling

First, you have to get and setup the necessary tooling. If you have already installed on your system Docker, docker-compose and Git tools, you can jump to Get the Sandbox :

• Follow the Get Docker CE for Ubuntu documentation
  • Do not forget the Post-installation steps for Linux
  • If you are behind a corporate proxy, follow Docker's instructions on how to configure your Docker daemon
• Install Docker Compose
• Install GIT, follow the Git Book instructions or if you are on a debian system, simply enter the command line : sudo apt-get install git

Get the IKATS Sandbox

1. Clone the ikats-sandbox GitHub repository

   mkdir ~/SCM
   cd ~/SCM
   git clone https://github.com/IKATS/ikats-sandbox.git

2. Create the necessary directories for persistent data (our docker_bindings)

   sudo mkdir -p /var/lib/ikats/IKATSDATA
   sudo chown -R ${USER}:${USER} /var/lib/ikats/

3. Get and extract the latest release of the file sandbox_docker_bindings.tar.gz (list of releases).
   Example, for release x.y.z:

   IKATS_RELEASE=x.y.z
   wget -P /tmp/ "https://github.com/IKATS/ikats-sandbox/releases/download/${IKATS_RELEASE}/sandbox_docker_bindings.tar.gz"
   tar xzf /tmp/sandbox_docker_bindings.tar.gz -C /var/lib/ikats
   rm /tmp/sandbox_docker_bindings.tar.gz

4. (optional) Get and extract data needed for IKATS tutorials

   IKATS_RELEASE=x.y.z
   wget -P /tmp/ "https://github.com/IKATS/ikats-sandbox/releases/download/${IKATS_RELEASE}/sandbox_hourly_weather_import_data.zip"
   mkdir /var/lib/ikats/IKATSDATA
   unzip /tmp/sandbox_hourly_weather_import_data.zip -d /var/lib/ikats/IKATSDATA
   rm /tmp/sandbox_hourly_weather_import_data.zip

Test your IKATS Sandbox

1. Launch the IKATS containers with Docker Compose

   cd ~/SCM/ikats-sandbox
   ./start_IKATS.sh

2. Depending on the .env file configuration, open to the IKATS Workbench in your browser http://${GUI_IP}:${GUI_PORT}

Next steps

• Read the Getting started tutorial
• You could import your "small" data. Remind that you are on a limited system.
• Other tutorials are available on the IKATS website starting from the How to... section
• Troubleshooting:
  • If you're comfortable with the command line, read the TROUBLESHOOTING.md page.
  • Feel free to contact the IKATS team via the support page
  • Feel a bug into the issues tracker

Release Notes

see changelog for details