Forward Historian Walkthrough

This guide describes a simple setup where one Volttron instance collects data from a dummy driver and sends it to another instance where it can be recorded.

I'm doing this example on a single machine. If you're doing the same I recommend the following:

• Set your Volttrons' vip addresses to use tcp. It's not needed but is closer to what a real deployment would have. If you're looking for open ports then $ netstat -ln --tcp will be helpful.
• Be careful with your terminal windows' environment variables. At least one of your Volttron instances won't be able to live in ~/.volttron
• If you run into trouble it can be helpful to run each agent in the foreground of its own terminal. This can be accomplished with the following:
  • Go to the agent's source directory
  • Locate the agent's configuration file and the python file with a main function (frequently called agent.py)
  • For the forwarding historian, the default config file is in the services/core/ForwardHistorian directory and its agent.py file is at services/core/ForwardHistorian/forwarder/agent.py
  • Use $ AGENT_CONFIG=config python -m forwarder.agent (no .py!) to run the forwarding historian in the foreground.

Configuration

1. Set up two Volttron instances as described in Deployment Walkthrough
2. In each Volttron's auth.json file add the public key and location of the other Volttron instance to the "allow" array.
3. Now the Volttron instances should be able to communicate. To avoid having to worry about authentication for the remainder of the walkthrough you can use --developer-mode when starting Volttron.

Forwarding Volttron

This Volttron will have a fake driver provided by the Master Driver and a Forwarding Historian. You may have to edit these configuration files before starting them:

services/core/ForwardHistorian/config
services/core/MasterDriverAgent/fake-msater-driver.agent
services/core/MasterDriverAgent/master_driver/test_fakedriver.config

Remote Volttron

This Volttron will receive the history sent from the Volttron instance we've already set up. Having a Listener Agent running in the foreground (scripts/launch_listener.sh) will make it easy to verify that the data has been successfully received.