About | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange

Running ROS across multiple machines

Description: This tutorial explains how to start a ROS system using two machines. It explains the use of ROS_MASTER_URI to configure multiple machines to use a single master.

Tutorial Level: INTERMEDIATE

Next Tutorial: Defining Custom Messages

Overview

ROS is designed with distributed computing in mind. A well-written node makes no assumptions about where in the network it runs, allowing computation to be relocated at run-time to match the available resources (there are exceptions; for example, a driver node that communicate with a piece of hardware must run on the machine to which the hardware is physically connected). Deploying a ROS system across multiple machines is easy. Keep the following things in mind:

• You only need one master. Select one machine to run it on.

• All nodes must be configured to use the same master, via ROS_MASTER_URI.

• There must be complete, bi-directional connectivity between all pairs of machines, on all ports (see ROS/NetworkSetup).

• Each machine must advertise itself by a name that all other machines can resolve (see ROS/NetworkSetup).

Talker / listener across two machines

Say we want to run a talker / listener system across two machines, named marvin and hal. These are the machines' hostnames, which means that these are the names by which you would address them when. E.g., to login to marvin, you would do:

ssh marvin

Same goes for hal.

Start the master

We need to select one machine to run the master; we'll go with hal. The first step is start the master:

ssh hal
roscore

Start the listener

Now we'll start a listener on hal, configuring ROS_MASTER_URI so that we use the master that was just started:

ssh hal
export ROS_MASTER_URI=http://hal:11311
rosrun rospy_tutorials listener.py

Start the talker

Next we'll start a talker on marvin, also configuring ROS_MASTER_URI so that the master on hal is used:

ssh marvin
export ROS_MASTER_URI=http://hal:11311
rosrun rospy_tutorials talker.py

Voila: you should now see the listener on hal receiving messages from the talker on marvin.

Note that the sequence of talker / listener startup doesn't matter; the nodes can be started in any order. The only requirement is that you start the master before starting any nodes.

Variation: connecting in the other direction

Now let's try it in the other direction. Leaving the master running on hal, kill the talker and listener, then bring them up on opposite machines.

First a listener on marvin:

ssh marvin
export ROS_MASTER_URI=http://hal:11311
rosrun rospy_tutorials listener.py

Now a talker on hal:

ssh hal
export ROS_MASTER_URI=http://hal:11311
rosrun rospy_tutorials talker.py

When something goes wrong

If something in the above sequence didn't work, the cause is likely in your network configuration. See ROS/NetworkSetup for configuration requirements and troubleshooting tips.