Dealing with Vlink multicast problems

Issue

Vlinks (virtual links) are aggregate bandwidth caps on transfers. Aspera uses multicast packets to communicate this information to servers.

If you run packet traces on your server segment, it may appear that there is too much multicast traffic. However, even a large amount of multicast traffic (which is usually at most 3-4 Mbps) does not make much of an impact your network.

In rare circumstances, this multicast traffic can cause problems with switches and routers in your network. You can take one of the two measures below depending on your network to alleviate this problem.

Multiserver networks

If you have more than one transfer server on your network, they will all need to see the Vlink multicast traffic being broadcast.

You can maintain this requirement while reducing the multicast traffic you see on the network by rerouting multicast traffic to a switch. The switch can be separate from the rest of the network entirely, or on the same switch within a separate VLAN.

Single server network

If you only have one server on the network and it’s on a Linux machine, you can redirect multicast traffic to the loopback port.

Create a file named route-lo in the following location:

  • /etc/sysconfig/network-scripts/route-lo

Edit the file to contain the following line, which redirects the multicast traffic to the loopback port:

239.255.0.0/24 via 127.0.0.1

It is necessary to restart the network service for this permanent change to take effect, which you can do with the following command:

# service network restart 
Have more questions? Submit a request

0 Comments

Article is closed for comments.
Powered by Zendesk