|NPACI Rocks Cluster Distribution: Users Guide:|
|Prev||Chapter 5. Customizing your Rocks Installation||Next|
For compute nodes, Rocks uses the first ethernet interface (eth0) for management (e.g., reinstallation), monitoring (e.g., Ganglia) and message passing (e.g., MPICH over ethernet). Often, compute nodes have more than one ethernet interface. This procedure describes how to configure them.
Additional ethernet interfaces are configured from the frontend via a command line utility named add-extra-nic. It manipulates the networks table on the frontend to add information about an extra interface on a node (a description of the networks table can be found in Rocks Cluster Schema) This program just manipulates the database. It does nothing to the actual configuration of a running node.
Once you have the information in the networks table, every time you reinstall, the additional NIC will be configured.
The structure supports multiple additional interfaces/node.
For each node that has an additional ethernet interface, execute:
# add-extra-nic --if=<interface> --ip=<ip address> --netmask=<netmask> --gateway=<gateway> --name=<host name> <compute node>
The name of the ethernet interface (e.g., eth1).
The internet address for the interface (e.g., 192.168.1.1).
The network mask for the interface (e.g., 255.255.255.0).
The gateway for this interface (e.g., 192.168.1.254).
Host name for the interface (e.g., fast-0-0).
The name of the compute node to apply the configuration to (e.g., compute-0-0).
For example, say you want to configure interface eth1 for compute node compute-0-0 with the IP address 192.168.1.1 with a netmask of 255.255.255.0 with a gateway of 192.168.1.254 and you want to name the new interface fast-0-0. The call to add-extra-nic would look like:
# add-extra-nic --if=eth1 --ip=192.168.1.1 --netmask=255.255.255.0 --gateway=192.168.1.254 --name=fast-0-0 compute-0-0
Reinstall the nodes that you have defined an additional interface (use shoot-node).