Red Hat Research Days Events 2021

A monthly series of virtual conversations between researchers and Red Hat experts

 

 

Watch the recordings from previous events

Connect with researchers and Red Hat experts to discuss research innovations in open source. Monthly research events include live video discussions where all attendees can join the conversations.

 

Join Upcoming Research Days Events

No event found!

Session Recordings 

Innovation Scorecard: Measuring the success of any innovation within business | September 22  

 

by Ondřej Žižlavský | Brno University of Technology, Eddie Fisher SKEMA Business School, BUT and Zdeněk Švécar | Red Hat

Show the abstract

The concept of innovation is not new. It relates closely to the concept of change. Both are inevitable in today’s and tomorrow’s business environments. Standing still and hoping for the best is no longer a viable option. Innovation, by itself, is not a panacea for positive accomplishments. Of paramount importance to any business is how successful any innovation has actually been.

This presentation brings together the knowledge, learning and experience from the authors’ practical applications of a newly developed and implemented Innovation Scorecard methodology to close exactly this shortfall. Our ‘one stop shop’ solution is a complete end to end approach on how to measure the success of any innovation, irrespective of whether this relates to projects or business as usual work environments. This landmark concept provides an applied proof of concept across a range of business applications and a complete end to end process ‘how to measure success’. The presentation offers a starter-pack to get you on the road to measuring the success of innovation. You will be provided with a simple to follow user-friendly process and the necessary tools and techniques to develop your own style to measure success within your business organisation.

Project website: iscorecard.org

Kubernetes Service Selection Optimization & Fine-grained Network Telemetry in Programmable Switches | July 21  

 

by Daniel Bachar | IDC Herzliya, Shir Landau-Feibish | Open University of Israel and Vishal Thapar | Red Hat

Show the abstract

Abstract 1: Kubernetes Service Selection Optimization
In the first talk, Daniel Bachar discusses his contributions to the Submariner project that provides the ability to connect multiple Kubernetes clusters such that services can communicate with each other. In such a multi-cluster environment, the same service may be deployed in different geo-locations, each with different cost and latency penalties. Yet, the current service selection mechanism does not take these locations and corresponding penalties into account. “I present a novel solution for optimizing this service selection. Interestingly, our solution utilizes Submariner’s Service Discovery process (Lighthouse) transparently. The simulations show our solution reduces the inbound traffic cost by up to 25% and the latency by up to 55%.”

Abstract 2: Fine-grained Network Telemetry in Programmable Switches

In the second talk, Shir Landau-Feibish discusses the drastically growing scale of today’s networks which makes managing them a significant challenge. Timely detection and response to events such as congestion, failure, and attack are crucial for proper network operation and require analyzing voluminous traffic quickly and accurately. To do so, we must devise new techniques for network monitoring and control that identify and fix problems when and where they happen. “I describe how programmable networks enable real-time network monitoring and specifically how we can detect queue buildup in programmable switches.”

Platypus: Power Side Channels In Software | June 23  

 

by Daniel Gruss, Moritz Lipp | Graz University of Technology and Jon Masters | Red Hat

Show the abstract

In this talk, we demonstrate how we can attack modern processors purely in software, using techniques from classical power side-channel attacks. We explain how we abuse the unprivileged access to energy-monitoring features of modern Intel and AMD CPUs. With Platypus, we show how to steal cryptographic keys from the operating system or trusted-execution environments, and how to break kernel address-space layout randomization within seconds. Finally, we discuss the mitigations that prevent our attacks.

A Plan for Practical Programming of FPGAs in the Data Center | April 22  

 

by Martin Herbordt | Boston University, Robert Munafo |  CAAD Lab at Boston University
and Ahmed Sanaullah | Red Hat

Show the abstract

To leverage the flexibility and performance potential of FPGAs in the data center requires either expensive specialized engineering talent, or commercial proprietary C-to-hardware tools that yield demonstrably poor performance. This is the performance portability programmability problem (P^4). 

In previous work, we found that there exists within current compilers the capability of delivering excellent FPGA performance for arbitrary C code, but that this capability is brittle, inconsistent, and requires some expertise on the part of the user to extract. Still, this result demonstrates that P^4 can be reduced to the problem of generating the correct sequence of optimizations for a particular input code and target architecture. Our hypothesis is that a solution to P^4 can be built using existing open source tools, primarily based on GCC, coupled with well-known machine learning techniques.

In this talk, we describe our plan in detail, together with problems to be solved, and outline our work to date. In particular, we report on an ongoing project that aims to use machine learning to control a newly customizable version of the GCC C compiler to automatically determine optimization pass ordering for FPGA targets specifically, and thereby improve performance as compared to existing (all proprietary) C-to-FPGA methods.

Mining Issued Common Criteria and FIPS 140-2 Certificates | March 24 

 

by Petr Švenda Masaryk University

Show the abstract

Security certification reports might be long, but they are also a trove of publicly available data about proprietary devices and other products otherwise available only under NDAs. While downloading and reading a single certificate is easy, reasoning about the characteristics of the whole ecosystem, which covers more than ten thousand certified devices based on human-written documents, is different. Are there observable systematic differences between the Common Criteria and FIPS 140-2 certificates? Can I quickly find out if my device is using a certified component recently found vulnerable? Most importantly, can we measure and quantify whether the whole process is actually increasing the security of the products being certificated? This talk address these questions using an open source tool for automatic analysis of publicly available certification reports, accompanied by catchy graphs.

The Future of Big Data: Massive Real-Time Data Streams | March 2  

 

by Ilya Kolchinsky | Red Hat, Technion
and Oren Oichman | Red Hat

Show the abstract

As we enter the era of Big Data, a large number of data-driven systems and applications have become an integral part of our daily lives, and this trend is accelerating dramatically. It is estimated that 1.7MB of data is created every second for every person on Earth, for a total of over 2.5 quintillion bytes of new data every day, reaching 163 zettabytes by 2025. In addition to the growing volume, velocity, and variety of continuously generated data, novel technological trends such as edge processing, IoT, 5G, and federated AI bring new requirements for faster processing and deeper, more computationally heavy data analysis. Therefore, we can no longer rely on the “old school” data processing mechanisms, and a new paradigm is needed.

In this talk, we explore this new paradigm, which we call Big Data Stream Processing (BDSP). BDSP comprises a variety of methods for scalable and efficient data processing that do not rely on traditional databases for storing and processing the data. We cover examples of real-life applications that can greatly benefit from incorporating BDSP capabilities.

In particular, we cover:
– On-the-fly detection of complex patterns in streaming data
– Stream-oriented machine learning and data mining

 

About Red Hat Research Days

Research Days events highlight Red Hat’s research initiatives. We bring together international researchers, Red Hat engineers, industry representatives, customers, partners, and others to share knowledge about the latest research findings, network with research enthusiasts from around the world and learn about moving great research ideas into open source communities.