Language Selector
| Español |
| Deutsch |
| 中文 (Greater China) |
| 日本語 (Japan) |
| 한국어 (Korea) |
Sharing ideas and data as well as equipment is crucial for scientists and researchers who are pushing the frontiers of physics, astronomy, biology, and other fields. Today, all science is “e-science,” and researchers need to be able to remotely access data generated by costly instruments, such as supercomputers or storage systems, and distributed computational resources, regardless of where in the world they reside.
Moreover, the need to collaborate with colleagues and access very large datasets in real time is creating demand for high performance networks. The global grid of computing, storage, and visualization resources is being extended to the LambdaGrid of computer resources connected by multi-lambda or multi Gigabit and 10 Gigabit networks. This new approach allows the networks themselves to be directly controlled resources – for example scheduled for dedicated use – just like the computer devices attached to them.
As a result, universities, research institutes, and government agencies are building a global e-science infrastructure comprised of computing, storage, visualization, and networking facilities, that enables researchers worldwide to share information, collaborate remotely, and integrate the resources they need to collect and process data. High performance networks are an essential component of this infrastructure, as are the exchange points that connect these international, national, regional, and local networks.
Since 2001, StarLight has been a cornerstone of the evolving e-science infrastructure. Located on Northwestern University’s downtown Chicago campus, StarLight is one of the world’s largest optical network exchanges for national and international research and education networks.

StarLight provides optical switching for wavelengths at one and 10 GbE speeds with its Gigabit Ethernet (GbE) switch/router facility. This advanced infrastructure also serves as a proving ground for network services optimized to support high performance, global e-science applications and experimental research equipment. To interconnect these varied networks, StarLight required the scalability and resiliency of the Force10 Networks TeraScale E-Series family of switch/routers.
Building a Powerful Exchange
With funding from the National Science Foundation, StarLight was developed by the Electronic Visualization Laboratory (EVL) at the University of Illinois at Chicago (UIC), the International Center for Advanced Internet Research (iCAIR) at Northwestern University, and the Mathematics and Computer Science Division at Argonne National Laboratory, in partnership with Canada’s CANARIE and the Netherlands’ SURFnet.
The networks that peer at StarLight include regional networks, state-wide networks (Illinois’ I-WIRE), national research and education networks (National LambdaRail and Internet2’s Abilene), all major federal agency networks (US Department of Energy’s [DOE] ESnet and UltraScience Network), international networks, and optical network testbeds (CAVEwave and OMNInet). Interconnecting these networks requires a highly scalable, highly reliable switch/ router capable of supporting large numbers of GbE and 10 GbE interfaces. After thoroughly testing and evaluating platforms, StarLight’s developers selected the Force10 Networks TeraScale E-Series to handle the demanding peering functions within the exchange.
“StarLight interconnects most of the world’s major high performance research and education networks in support of large scale, data-intensive e-science applications and requires not only scalable high performance networking but also resiliency and reliability,” said Tom DeFanti, director of the University of Illinois at Chicago’s Electronic Visualization Laboratory and one of StarLight’s founders.
“The Force10 E-Series enables us to provide a foundation upon which to build large scale, cost-effective network services without compromising performance, especially as additional 10 GigE circuits are added and data volumes increase.”
The Force10 E-Series combines leading port density and performance with high availability and resiliency features. The E1200, for example, supports up to 1,260 GbE and 224 10 GbE ports. Currently, StarLight has nearly 36 line-rate 10 GbE ports and 96 one GbE ports in use as well as additional ports for testbed research. The E-Series line cards, switch fabric, backplane, central processor, and operating system have all been optimized to process terabits of traffic at line rate in a reliable, predictable fashion.
Resiliency features include the E-Series three- CPU architecture, which distributes switching, routing, and management functionality between distinct processors. This design allows faults to be contained while protecting other parts of the system. In addition, the E-Series has a fully redundant architecture, including fault-tolerant switch fabrics with hitless failover technology that ensures the secondary components are activated with zero packet loss in the event of a failure.
The port density, availability and resiliency features of the Force10 E1200 provide a scalable, cost-effective foundation for StarLight’s high performance exchange. The high scalability of the E-Series means that a single chassis can deliver the functionality that other vendors would need multiple boxes to provide, which translates to lower capital, operational and management expenses.
While many of the advanced networks connected to StarLight support production science, several links are used exclusively for LambdaGrid application experiments. StarLight research is developing new architectures, technologies, and techniques that allow for unprecedented levels of distributed network resource integration. Among the methods used are optical signaling, control and management protocols, new UDPbased transport protocols, security and authentication capabilities, data mining techniques for real-time correlation, and real-time visualization and collaboration toolkits for very-large datasets.

Collectively, these advanced networks comprise the Global Lambda Integrated Facility (GLIF), an international partnership of countries, consortia, institutions, and individual research initiatives that support data-intensive scientific research and middleware development.
“The Force10 equipment is currently supporting multiple national and global advanced applications, including leading edge prototypes that cannot be supported by traditional infrastructure,” said Joe Mambretti, director of iCAIR and manager of the StarLight facility.
As one of the platforms for StarLight, the E1200 is helping support the growing e-science infrastructure that will enable researchers to push the frontiers of science even as StarLight’s developers continue to push the boundaries of technology.
StarLight is an advanced optical infrastructure and proving ground for network services optimized for high performance applications. Operational since summer 2001, StarLight is a one GbE and 10 GbE switch/router facility for high performance access to participating networks and also offers true optical switching for wavelengths. StarLight is being developed by the Electronic Visualization Laboratory (EVL) at the University of Illinois at Chicago (UIC), the International Center for Advanced Internet Research (iCAIR) at Northwestern University, and the Mathematics and Computer Science Division at Argonne National Laboratory, in partnership with Canada’s CANARIE and the Netherlands’ SURFnet. STAR TAP and StarLight are made possible by major funding from the US National Science Foundation to UIC. StarLight is a service mark of the Board of Trustees of the University of Illinois. See www.startap.net/starlight.

GLIF is a collaboration of institutions, organizations, consortia, and country National Research and Education Networks (NRENs) that voluntarily share optical networking resources and expertise for the advancement of scientific collaboration and discovery. GLIF’s mission is to create and sustain a global facility that supports leadingedge capabilities based on new and emerging technologies and paradigms related to advanced optical networking. These capabilities will enable high-performance applications and services, including the timely transfer of massive amounts of data, distributed computing, data analysis, collaboration and visualization, and control of remote instruments. GLIF provides leadership in advanced technologies and pre-production services on behalf of NRENs, NREN consortia, or pan-continental R&E networks, creating new models that they can implement. Once such services are available from NREN consortia, GLIF will refocus on new and emerging paradigms to support its communities. For more information, see www.glif.is.