Microsoft and University of California, San Diego (UCSD) researchers have separately developed divergent methods for solving shortcomings in data center architectures. Microsoft aims to provide high performance for all traffic, regardless of demand, while the UCSD researchers focus on allowing the free migration of virtual machines (VMs), minimal configuration when adding new hosts to a network, and quickly solving failures. Microsoft’s approach also addresses VM migration and Layer 2-like addressing, but uses a method that calls for installing an agent on every endpoint, which conflicts with UCSD’s approach to adjust switch software and leave the endpoints alone. The UCSD researchers, led by professor Amin Vahdat, propose blending Layer 2 and Layer 3 connectivity in data centers, which would allow for massive scaling that is otherwise limited by Layer 2 factors and reduces the management and configuration demands of Layer 3. UCSD’s PortLand scheme would support VM migration, which Layer 3 cannot do because VMs can move from server to server, each with a different IP address. Microsoft’s VL2 architecture handles the addressing problem by introducing a two-tiered system: a location-specific IP address, and an application-specific IP address that follows applications as they move between VMs. In the VL2 architecture, each server is associated with the location-specific IP address of the switch it is attached to and a VL2 directory system maps the location IPs to the application IPs. VL2 has some features PortLand lacks, including the ability to refuse to provide the location-specific IP address if access policies deny the initiating server connectivity to the destination server. The researchers presented their work at ACM’ SIGCOMM 2009 conference, which takes place Aug. 17-21 in Barcelona, Spain.
For More Information:http://www.cpccci.com
