Versions Compared


  • This line was added.
  • This line was removed.
  • Formatting was changed.

In a typical data center, connections between servers and the leaf or top of rack switches are often done at layer 2. In order to build more resilient data centers, many Cumulus Networks customers are leveraging the Linux ecosystem to run routing protocols directly to their servers, running layer 3 protocols like OSPF (Open Shortest Path First) or BGP (Border Gateway Protocol) directly on the hosts, so the hosts can participate directly in the routing fabric. This is often referred to as Routing on the Host.

In Cumulus Linux 3.0 and later, Routing on the Host works on server hosts in a number of different environments:


Cumulus Networks enhanced Quagga with the ability to implement RFC 5549, so you can configure BGP or OSPF unnumbered interfaces on Cumulus Linux switches, and now, with Routing on the Host, on server hosts and containers as well. In addition to the benefits of not having to configure every subnet described above, you do not have to configure anything specific on the leaf switch at all, so you don't have to configure an IPv4 address in /etc/network/interfaces for peering.


Another benefit of Routing on the Host is the ability to gracefully remove a leaf switch from the fabric for maintenance. With layer 2 only (like MLAG), you cannot influence routes without being disruptive (that is, some traffic loss must occur). With OSPF and BGP, there are multiple load balanced routes via ECMP (Equal Cost Multipath) routing. Since there is routing, it is possible to change these routes dynamically.