NCSA Provides Update on FABRIC Research Infrastructure Project
May 16, 2022 – When the Internet was launched in the last century, it was an incredible feat of science and technology. But it was not designed for the massive datasets, machine learning tools, advanced sensors and Internet of Things devices that have become central to many research and business projects and our homes.
Funded with a $20 million grant of national science foundation, FABRIC (Adaptive Programmable Research Infrastructure for Computer Science and Science Applications) explores ways to replace an Internet infrastructure that has been aging for 20 years.
The NCSA is one of 13 institutions collaborating to create a platform for testing new Internet architectures that can enable a faster, more secure Internet that is better suited to today’s users and future needs. One who is also able to do impossible things now. The FABRIC project is led by the Research and Network Infrastructure Group to RENCI at the University of North Carolina at Chapel Hill.
NCSA: First with concerts
Last fall, NCSA installed a 100 gigabit network connection dedicated solely to the FABRIC projectthe first FABRIC employee to do so.
The NCSA already had six 100-gigabit Internet connections, says David Wheeler, manager of NCSA’s ICI data delivery and management division and principal investigator of NCSA’s FABRIC node. The FABRIC connection, however, is used only for research on FABRIC’s purposes and is connected to FABRIC’s 400 gigabit backbone.
Wheeler says the NCSA Networking Team supports FABRIC design in NCSA environments, hosting a FABRIC skein (networking node) and an experimenter-oriented storage system.
NCSA network engineer Corey Eichelberger installed the rack, coordinated connections, and ensured that the FABRIC equipment was isolated from other NCSA network nodes. The isolation is partly for safety but mainly to allow full accessibility of the test bench. Matt Kollross, head of the network engineering team, and network engineers Eric Boyer and Michael Douglas also contribute their expertise.
“People don’t often think much about a network. Networking is a given from many people’s perspective. But I think those are the fascinating underpinnings of what we do. Science is increasingly relying on the Internet. This need will only grow. And the internet we have now won’t be able to keep up, said David Wheeler, manager of NCSA’s data management and delivery division HERE.
He notes that it is the work of network engineers that enables people to have and use the Internet, whether for personal or professional use or for scientific research.
Research intensification
This is the third year of the four-year FABRIC project and this is the point “where it gets really exciting,” says Anita Nikolich, research scientist and director of research innovation at the I’m at school at the University of Illinois Urbana Champaign and co-PI on the FABRIC project.
This is the phase where employees can experiment and test. It’s a networking lab. Although the NCSA grant and investment in the project is modest, approximately $170,000, being able to offer networking insight, expertise, and a safe environment for scholars to be creative and try to new things is priceless.
“FABRIC’s intention,” says Nikolich, “is a place where people can [try new things] and experiment to build the new Internet, one that is secure, censorship-resistant, and supportive of big science. More than just a big network, a network that really supports science.
What NCSA is looking for
While FABRIC focuses on the United States, the science is conducted globally. Thus, Nikolich is the principal investigator of an additional grant to focus on the international appeal FABRIC beyond borders. FAB is an extension of the FABRIC test bed linking North America’s central infrastructure with four nodes in Asia, Europe and eventually South America, creating the networks needed for science to move large amounts of data across oceans and time zones seamlessly and securely.
A key contributor to FAB, says Nikolich, is the NCSA team working on Investigating the Rubin Observatory’s Space and Time Legacy.
“Basically, FABRIC will serve as a test bed to try to get data streams into the hands of scientists faster. And once it has proven itself on FABRIC, LSST will ideally include it in its scientific workflow,” says Nikolich.
The NCSA LSST team is experimenting with adjustments to the software used to transmit data to data brokers. The focus is on optimizing alert feeds so that when the telescope becomes operational in Chile later this year, scientists will receive alerts faster than they do now.
As FAB currently has no nodes in South America, the team is emulating what a network connection would be like. They begin by taking a long-lasting slice of Docker and emulate processing to operate test alert streams. (Docker is a type of container, an application that bundles code and dependencies together so software always runs the same way, regardless of infrastructure.) The next step will be to emulate the move from Chile to Node FABRIC here.
“We want to make sure they test how to do this particular networking angle,” says Nikolich, “so that when the actual network is installed in the next couple of years, they already have the optimized software. FABRIC helps prepare for this workflow. »
An AI Internet that is also secure
According to Nikolich, some other research areas that various FABRIC collaborators focus on are cybersecurity, censorship evasion, and artificial intelligence, commonly known as AI.
Censorship evasion
For example, she says, a Clemson experimenter is exploring how to make sure the future internet is censorship-resistant so that there is free and open communication. This is an important societal issue, she notes, as we have recently seen entire countries disconnected from the internet. Networking is crucial in providing barriers to censorship.
AI innovations
Advanced chip functionality in FABRIC nodes has led to AI experiments and more experiments are planned in the coming months. These include network processing and adversarial AI.
Network processing, Nikolich explains, is exactly what it sounds like. In traditional networks, packets – small amounts of data sent over a network – are reassembled into a single file or other contiguous block of data at the receiving end.
With network processing, the software processes data as it is transmitted over the network, removing things or doing things to add value or doing analytics or doing machine learning . Nikolich says physicists at CERN are experimenting with network processing, transmitting data from CERN to Chicago and the NCSA. “It’s just more efficient,” she notes.
“For scientists, that’s huge,” says Nikolich. “They don’t have to ship a ton of data and wait. They can do it along the way.
With two AI institutes in Illinois, Nikolich hopes to do more AI-focused experiments to show the power of networked processing for AI.
Adversarial AI
Encouraging people to secure the next Internet with the future in mind is also part of the FABRIC process, for example, Nikolich and Wheeler. Even though it is open science, how do you protect it? What can an opponent do? How to protect yourself from it?
This is how the Internet of the future will be more secure, they say. Thinking like an adversary now, not 30 years from now.
The fabric test bench
FABRIC consists of storage, compute and network hardware nodes connected by dedicated high-speed optical links. All major aspects of the infrastructure are programmable so researchers can create new configurations or adapt the platform for specific research purposes, such as cybersecurity.
Central nodes are deeply programmable, meaning they are controlled by network owners and can be programmed according to their needs. This flexibility and control over network functionality at all network points allows experimenters to test new architectures not possible today.
The FABRIC network is also expandable, meaning it is designed to allow for the addition of new capabilities, facilities and features such as cloud, networking, other testbeds, computing facilities and scientific instruments.
It looks like the internet of the future.
Source: Barbara Jewett, NCSA