Analyzing security assertions in High-Performance Computing environments

worked on by:Abraham Söyler

Outline/Expose

Security assertions are an important factor when deciding on the feasability of
computer systems. While there has been much research on security in general
purpose and personal machines, high-performance computing (HPC) systems have
inherently different security properties. Most of the research effort for HPC
systems has been focused on performance issues and disregarded security aspects
as less relevant.

Yurcik et al. (2004) identified that emergent properties are key considerations
of cluster security and outlined methods and tools for monitoring. They also
provide a set of research challenges around cluster security in 2005.

The Partnership for Advanced Computing in Europe (PRACE) published several works
regarding HPC systems. Nowak et al. (2013) based their work on a survey in the
PRACE community, providing insight in the used security technologies and
assertions and giving general recommendations on the security of HPC
infrastructure. Peisert (2017) explored the security differences in HPC to
typical IT contexts and the challenges involved in securing them.

In this work we will summarize the current state of research on security in HPC
systems, their assertions and proposed methods. Furthermore we shall explore
standardization efforts made by the BSI and ISO relevant for security aspects in the creation,
management and usage of HPC systems and propose a method of evaluating the
security of such environments.

[1] Yurcik, William et al. “Cluster Security as a Unique Problem with Emergent Properties: Issues and Techniques.” (2004).
[2] Michał Nowak. (2013). Security in HPC Centres.
[3] Peisert, Sean (2017). "Security in High-Performance Computing Environments". Communications of the ACM, September 2017, Vol. 60 No. 9, Pages 72-80

Thesis Requirements

formulate requirements here (together with your adviser)

Milestones and Planning

Past CW Goals
DONE CW7 Exposè
  CW9 introductory presentation
  CW11 collection and review of relevant research with supervisor
  CW13 first draft of evaluation method and setup/architectural structure of exemplary HPC environment
  CW15 application of evaluation method to exemplary environment
  CW16 finished thesis
  CW18 feedback and presentation

Weekly Status

Week 1 (CW 6)

Activities
  • Exposè written
  • gathered first bunch of research papers

Results

  • Exposè

Next Steps

  • review research papers and sort for relevance; summarize
  • prepare git repository for thesis and resources

Problems

  • research is quite sparse for this topic