RSDA 2021

KEYNOTES

We are pleased to announce the RSDA 2021 Keynotes by:

Karthik Pattabiraman
Professor of Electrical and Computer Engineering (ECE) at the University of British Columbia in Canada

Title:
Failure Data Everywhere, but Not a Failure in Sight! The joys and frustrations of analyzing open data sources

Abstract:
The conventional wisdom is that reliability (i.e., failure) data is difficult to obtain, and even where it is available, there are many restrictions on its use (e.g., NDAs). In this talk, I will argue that reliability data is actually quite easy to obtain, provided one is willing to look past the conventional sources. This also means being flexible with the definition of a failure, and being willing to accept some messiness and chaos in the data, as ground truth is often not available.

I will share some anecdotes from my research over the last decade, where going beyond the conventional sources of failure data ended up paying rich dividends. These were data sources that were either publicly available or easily accessible without any restrictions. However, we had to stretch our definition of what a failure is, and also reconcile conflicting or missing information. These data sources span the domains of the world wide web, cloud data centers, machine learning workloads, and Ethereum smart contracts. I will conclude by presenting some of the lessons learned, and future challenges. This is joint work with my students and collaborators.

Massimiliano Rak
Associate professor at University of Campania Luigi Vanvitelli

Title:
The Need for Security Automation

Abstract:
Nowadays, everyone speaks about cybersecurity. As a fact, the number of cybersecurity incidents has grown a lot in the last years. Probably as a side effect the widespread diffusion of smart systems, lived the pervasiveness of Internet of things and of home automation system together with vocal assistants, and last but not least, of the incredible grow of adoption of smart working practice, with the covid-19 pandemic. Or, together with the above cited factors, due to the fact that security practices are pretty far to be commonly adopted. Security has a cost, but it does not have a measurable direct return of such expenses (Well… missing security MAY cost a lot more then security practices, but you will be aware of this only when it is too late…). Adoption of practices that automate security processes and facilitate their adoption could be the only way to address such issues. But …. Is security automation possible? How much we can automate it? Are such automation compatible with the new development methodologies and practices (Agile, Devops, Continuous Practices, …)?

KEYNOTE SPEAKERS IN PREVIOUS RSDA EDITIONS:

• RSDA 2020
  Saurabh Bagchi, Purdue University: Dependability Meets Data Analytics

• RSDA 2019
  Ingo Weber, TU Berlin: Behavioural Analytics and Blockchain Applications - a Reliability View

• RSDA 2016
  Hervé Debar, Télécom SudParis: Towards a quantitative approach for threat mitigation and response

• RSDA 2014
  Leyla Yumer, Symantec Research Labs: From Script-kiddies to Cyberwars...
  Brendan Murphy, Microsoft Research: The Role of Data Analytics in Reliability and Security Verification

• RSDA 2013
  Nuno Silva, Critical Software: The Role of Data for Safety Critical Systems Development and Validation