Skip to content. | Skip to navigation

Personal tools

Navigation

You are here: Home / Wiki / Kb88

Kb88

Emulab FAQ: Hardware Setup: What is the control net and how do I identify it on a node?

Emulab FAQ: Hardware Setup: What is the control net and how do I identify it on a node?

What is the "control network"?

All experimental nodes in the testbed are connected to a common LAN called the control network. The control network is used for a variety of purposes including: loading node disks, allowing nodes to communicate with the Emulab infrastructure, and allowing access to the nodes from outside. For a more complete description see the tutorial section.

The fact that it is a shared LAN has many implications for applications, including for traffic shaping and routing. All stem from the fact that network traffic can traverse the control network between experiment nodes rather than the designated topology. Again, refer to the tutorial for more details.

How do I identify the control network interface on a node?

Since different types of nodes may have different control net interfaces and different OSes identify the interfaces in different ways, you cannot just hardwire a particular name in your scripts.

Instead, use the file /var/emulab/boot/controlif which, on any experiment node, contains the name of the control network interface. For example, you could do:

tcpdump -i `cat /var/emulab/boot/controlif`