Communications of the ACM
Computer Scientists Scale 'Layer 2' Data Center Networks to 100,000 Ports and Beyond
Credit: iStockphoto
University of California, San Diego (UCSD) researchers have developed software designed to enable data centers to function as single, scalable, plug-and-play networks. The PortLand software system is a fault-tolerant, layer 2 data center network fabric capable of scaling to more than 100,000 nodes.
The researchers say the software could help support large-scale data center networks by increasing inherent scalability, providing baseline support for virtual machines and migration, and significantly reducing administrative costs. PortLand also removes the reliance on a single spanning tree and can natively leverage multipath routing and improve fault tolerance.
"With PortLand, we came up with a set of algorithms and protocols that combine the best of layer 2 and layer 3 network fabrics," says UCSD professor Amin Vahdat. "Ideally, we would like to have the flexibility to run any application on any server while minimizing the amount of required network configuration and state."
One of PortLand's key innovations is its location discovery protocol, which allows for the possibility of a scalable layer 2 network. Switches automatically learn their location within the data center topology without human assistance, and can then assign Pseudo MAC addresses to each of the servers they connect to. "The goal is to create a network fabric that allows you to buy any server or switch, plug it in, and have it just work," says UCSD computer science graduate student Radhika Niranjan Mysore. The PortLand system was presented at ACM's SIGCOMM 2009 conference, which takes place August 17–21 in Barcelona, Spain.
From UCSD News
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA 
No entries found