2022 Technical Program (June 3)

Technical Program [June 3, 10:00 AM - 5:30 PM ET]

Workshop Opening [10:00 AM - 10:05 AM ET]

• Workshop Opening and Welcome
  Julita Corbalán, Gonzalo P. Rodrigo and Dalibor Klusáček

Technical Papers (session 1) [10:05 AM - 12:00 AM ET]

• Toward Building a Digital Twin of Job Scheduling and Power Management on an HPC System.
  Tatsuyoshi Ohmura, Yoichi Shimomura, Ryusuke Egawa and Hiroyuki Takizawa
• Dynamic Management of CPU Resources Towards Energy Efficient and Profitable Datacentre Operation .
  Christos Kalogirou, Christos D. Antonopoulos, Spyros Lalis and Nikolaos Bellas
• Improving Accuracy of Walltime Estimates in PBS Professional using Soft Walltimes .
  Dalibor Klusacek and Vaclav Chlumsky
• Optimization of Execution Parameters of Moldable Ultrasound Workflows under Incomplete Performance Data .
  Marta Jaros and Jiří Jaroš

Coffee Break & Discussion [12:00 AM - 12:15 AM ET]

Technical & OSP Papers (session 2) [12:15 AM - 1:45 PM ET]

• AI-Job Scheduling on Systems with Renewable Power Sources .
  Ganesh Kamath Nileshwar and Uwe Schwiegelshohn
• Using Kubernetes in Academic Environment: Problems and Approaches (Open Scheduling Problem) .
  Viktoria Spisakova, Dalibor Klusacek and Lukas Hejtmanek
• Re-making the Movie-making Machine.
  James Vanns and David Galeano

Keynote Lecture [1:45 PM - 2:45 PM ET]

• Insights and Requirements for Future Workflow Scheduling
  Lavanya Ramakrishnan (Lawrence Berkeley National Lab, USA).
  
  Abstract: Workflows are an important construct in scientific ecosystems that capture the relation between different steps of processing and data. Workflow tools have focused on providing automation and repeatability but mostly consider HPC resources as blackboxes. In this talk, I will outline the evolution of scientific workflow needs over the last 20 years and its impact on workflow scheduling on HPC systems. I will outline a) real science use cases with challenging scheduling problems, b) capture our analyses of workloads on HPC systems over four different large-scale systems, c) discuss the implications of data, resources, networks, containers, interactive notebooks on scheduling. I will conclude the talk by discussing the future challenges of scientific workflow scheduling that supports autonomous experimentation and observation in concert with a self-driving infrastructure.
  
  Biography: Lavanya Ramakrishnan is a Senior Scientist and Division Deputy in the Scientific Data Division at Lawrence Berkeley National Lab. Her research interests are in building software tools for computational and data-intensive science with a focus on workflow, resource, and data management. More recently, she has been using user research methods to understand as well as verify/validate the context of use and social challenges that often impact tool design and development. She currently leads several project teams that consist of a mix of social scientists, software engineers, and computer scientists.

Coffee Break & Discussion [2:45 PM - 3:00 PM ET]

Technical Papers (session 3) [3:00 PM - 5:30 PM ET]

• Encoding for Reinforcement Learning Driven Scheduling .
  Boyang Li, Yuping Fan, Michael Papka and Zhiling Lan
• Scheduling of Elastic Message Passing Applications on HPC Systems .
  Debolina Halder Lina, Sheikh Ghafoor and Thomas Hines
• RARE: Renewable Energy Aware Resource Management in Datacenters .
  Vanamala Venkataswamy, Jake Grigsby, Andrew Grimshaw and Yanjun Qi
• RADICAL-Pilot and PMIx/PRRTE: Executing heterogeneous workloads at large scale on partitioned HPC resources .
  Mikhail Titov, Matteo Turilli, Andre Merzky, Thomas Naughton, Wael Elwasif and Shantenu Jha
• On the Feasibility of Simulation-driven Portfolio Scheduling for Cyberinfrastructure Runtime Systems.
  Henri Casanova, Yick Ching Wong, Loic Pottier and Rafael Ferreira da Silva