Skip to content

2. Setting up worker nodes

Jump to bottom
Brice Nichols edited this page Mar 3, 2016 · 2 revisions

Open Communication to the Web App

To set up servers for running models on demand from the web app:

  • clone the queue repository
  • install [Anaconda Python] (https://www.continuum.io/downloads) 2.7 if you don't already have it
  • install [git command line tools] (https://git-scm.com/downloads)
  • pip install:
    • Pyro4
  • copy "node.py" to a directory where we want model runs stored
  • run node.py from the shell: python node.py
    • note: the shell may need to be run as a administrator
  • this sets up a communication port on the node worker machine, waiting patiently for a command from the web-app

Preparing for Model Runs

Each worker node should be configured and ready to run the model that will be requested from the web application. Install all software needed and configure the machine to run your model, as if it was a standard workstation. Make sure that it works in "standalone" mode first before moving ahead. For instance, if the node is to be used to run Soundcast, follow the [setup guidelines for running Soundcast] (http://soundcast.readthedocs.org/en/dev/).

Clone this wiki locally