IC2E 2023 Program

All session times refer to Eastern Daylight Time (EDT/GMT-4).

9:00 AM

Practical Reproducibility Tutorial
Kate Keahey and Ibrahim Matta

BU Center for Computing and Data Sciences, 665 Commonwealth Avenue, Boston, MA 02215

Reproducibility is not just a sound scientific practice – it also has many potential practical applications in teaching and creating new research results. In this tutorial, the attendees will learn about how to package their computer science research experiments in such a way that they are not only reproducible – but also practically reproducible, i.e., capable of being reproduced easily enough to be used as a mainstream method of interactive scientific exploration and exchange.

To achieve this objective this tutorial will be part instructional and part hackathon. The instructional part will cover how to use the NSF-funded Chameleon platform, one of the largest academic clouds, and how to package experiments for reproducibility using this platform. Attendees will learn about tools and services Chameleon provides to share experiments, including using Jupyter to manage the full experimental workflow including creating the environment, implementing experiment body, and data analytics; using Chameleon daypass to give access to the testbed for reproducibility; as well as Trovi, an experiment sharing portal integrated with the testbed. The tutorial will also cover the existing experiment patterns available via Trovi representing common elements of experimental configurations such as e.g., configuring storage with RAID, NFS, or RDMA.

The second part of the tutorial will consist of a hackathon in which the attendees will either reproduce existing experiments from a list of recommendations, or participate in a “reproducibility clinic” allowing them to partially or fully package their own experiments with the help of the organizing team. After the event, participants can share their packaged experiments with their own community so that other researchers can reproduce their experiments (and potentially cite their work).

9:00 AM

New England Research Cloud Tutorial (1/2 day)
Milson Munakami

BU Center for Computing and Data Sciences, 665 Commonwealth Avenue, Boston, MA 02215

In today's rapidly evolving digital landscape, organizations and researchers are constantly seeking ways to harness the agility and scalability of cloud computing while maintaining control and security over their data. The vision of the New England Research Cloud (NERC) is to offer a regional resource, rich with suites of on-premise cloud computing services.

Join us for a comprehensive tutorial workshop that will demystify NERC offered on-premise cloud services and provide you with the knowledge and practical skills to leverage this technology effectively. Whether you are an IT professional, a researcher, or a business leader, this workshop is designed to empower you with the tools and expertise needed to deploy and manage on-premise cloud services within your institution.

Tutorial Highlights:
  • Introduction to NERC and its core services: Gain a solid understanding of the vision behind the NERC core services and why they are essential in today's computing landscape.
  • Use Cases and Case Studies: Explore real-world scenarios and success stories where academic institutions and researchers have adopted NERC for their research endeavors.
  • Hands-on Demos:
    1. Running LLAMA-2 on NERC OpenStack: How to run Meta's LLAMA-2 Large Language Model (LLM) chat model on a GPU based virtual machine running in a Jupyter notebook.
    2. Integrating GitOps on NERC OpenShift: How to setup Cloud Native CI/CD Pipeline connected to the NERC’s OpenShift project.
    3. MLOps on NERC's Red Hat OpenShift Data Science (RHODS): How to have end-to-end machine learning workflows run on NERC’s RHODS setup.
    4. Q&A and Networking: Engage in interactive Q&A sessions to share insights and experiences.

By the end of the tutorial, you will have a thorough understanding of NERC, on-boarding process, and its offered core services, and will also be able to apply this knowledge to your own research projects.

Join us for this immersive learning experience and embark on a journey to unlock the power of NERC services.

1:30 PM

OCT FPGA Workshop (1/2 day)
Suranga Handagala

BU Center for Computing and Data Sciences, 665 Commonwealth Avenue, Boston, MA 02215

The Open Cloud Testbed (OCT) is a CISE Community Research Infrastructure (CCRI) project, supported by NSF at the Grand level, that provides researchers access to network-connected FPGA- enhanced server nodes through the CloudLab framework. CloudLab nodes are bare metal, meaning they are provided without an operating system or any pre-installed software or tools. This provides users with a flexible and powerful computing environment that can be customized to meet their unique needs. A distinctive feature of OCT is the exposure of its network interfaces to users, a capability not typically available on commercial and private clouds. This feature enables users to conduct advanced research using network-attached FPGAs, taking advantage of a high-speed network, and making new discoveries in various research fields.

This tutorial will guide participants through the process of building and deploying FPGA applications in the OCT, providing an overview of the FPGA development tools required to create bitstreams and deploy them on FPGA hardware. Participants will also gain insight into the applications and benefits of network-attached FPGAs. The tutorial will cover the following topics:

  • Introduction to OCT and its FPGA capabilities
  • Overview of the FPGA development process in OCT
  • Building and deploying FPGA-based applications in OCT
  • An example of using network-attached FPGAs
  • Research directions for network-attached FPGAs
By the end of the tutorial, participants will have a thorough understanding of the FPGA build and deployment process in the OCT, and will also be able to apply this knowledge to their own research projects. This tutorial is ideal for researchers who are interested in exploring the potential of FPGA-enabled computing in various data center and cloud computing applications.

Doctoral Symposium

The Doctoral Symposium has been moved to Tuesday, September 26, at 2:30 pm!

8:15 AM - 8:30 AM

8:30 AM - 9:30 AM

Keynote: Scaling Over the Bounded Sea of Data and Jobs: Towards a Holistic Approach
Larry Rudolph (Two Sigma Investments LLP and Mass Open Cloud, USA)
Keynote chair: Orran Krieger

43 Hawes Street, Brookline, MA 02446 Although both data and jobs continue to scale, it is important to consider them together along with the needs of tenants in and providers of clouds. It is not only the size of data sets that is scaling but the number of data sets, their complexity, and the number of data scientist interacting with each data set. The challenge is mostly dependency management but, unfortunately SQL often does not expose the dynamic dependencies between data sets or their components. In terms of jobs, it is the web of data pipelines that is challenging but, focus on the job or service level often does not expose the lineage and metadata necessary for pipeline optimization. Similarly, cloud tenants may know a lot about their jobs and data sets but little about the just a bunch of storage and compute servers while cloud providers know a lot about their storage and servers but little about the just a bunch of jobs and data sets. More research is needed in holistic optimization that combine all of these by share domain knowledge.

9:30 AM - 10:00 AM

Coffee break

10:00 AM - 12:00 PM

Session 1: Containers and micro-services
Session chair: David Bermbach

  • Efficient Container Image Updating in Low-bandwidth Networks with Delta Encoding (Research Track)
    Naoki Matsumoto, Daisuke Kotani, and Yasuo Okabe
  • Protocol-Independent Context Propagation for Sharing Microservices in Multiple Environments (Research Track)
    Hiroya Onoe, Daisuke Kotani, and Yasuo Okabe
  • Enriching Cloud-native Applications with Sustainability Features (Research Track)
    Monica Vitali, Paul Schmiedmayer, and Valentin Bootz
  • Breaking the Vicious Circle: Self-Adaptive Microservice Circuit Breaking and Retry (Research Track)
    Mohammad Reza Saleh Sedghpour, David Garlan, Bradley Schmerl, Cristian Klein, and Johan Tordsson

12:00 PM - 1:30 PM

Lunch

1:30 PM - 2:30 PM

Keynote: Predictability in the Edge-to-Cloud Computing Continuum
Alessandro Papadopoulos (Mälardalen University, Sweden)
Keynote chair: Stefan Schulte

43 Hawes Street, Brookline, MA 02446 Edge and Fog Computing are new computing layers between data-generating devices and the cloud, leading to the emergence of an edge-to-cloud computing continuum. Their role is to complement Cloud Computing pushing the intelligence outside of the devices but closer to the network edge. Applications with strict timing requirements can significantly benefit from offloading some of their functionalities toward the computing continuum. In turn, the computing continuum is expected to deliver a predictable behavior, in terms of deterministic data processing, minimal latency, and real-time responses. This keynote will discuss challenges and solutions for the development of predictable components in the computing continuum.

2:30 PM - 4:00 PM

Coffee break and Poster Session

2:30 PM

Doctoral Symposium

4:00 PM - 6:00 PM

Session 2: Energy
Session chair: Monica Vitali

  • A Framework for Enabling Cloud Services to Leverage Energy Data (Research Track)
    Vasileios Karagiannis, Shievam Kashyap, Nikolas Zechner, Oliver Hödl, Georg Hartner, Manuel Llorca, Tooraj Jamasb, Stefan Grünberger, Marc Kurz, Christoph Schaffer, and Stefan Schulte
  • Energy-efficient OECT Sensor Data Analysis on Constrained Edge Devices (Research Track)
    Francesco Saccani, Manuele Bettelli, Francesco Gentile, and Michele Amoretti
  • Towards Serverless Sky Computing: An Investigation on Global Workload Distribution to Mitigate Carbon Intensity, Network Latency, and Cost (Research Track)
    Robert Cordingly, Jasleen Kaur, Divyansh Dwivedi, and Wes Lloyd
  • GreenCoin: A Renewable Energy-Aware Cryptocurrency (Research Track)
    Nazmus Saquib, Shivaansh Kapoor, Chandra Krintz, Rich Wolski, and Markus Mock

7:00 PM - 9:00 PM

8:30 AM - 9:30 AM

Keynote: The Open Cloud and its Impact on Research
Orran Krieger (Boston University, USA)
Keynote chair: Michael Zink

BU Center for Computing and Data Sciences, 665 Commonwealth Avenue, Boston, MA 02215 While Cloud Computing is transforming society, today's public clouds are black boxes, implemented and operated by a single provider that makes all business and technology decisions. The Mass Open Cloud (MOC) was launched in 2013 with the vision of creating a production cloud that would enable innovation by a broad industry and research community. Today, the MOC Alliance provides a structure for a set of interrelated projects including production cloud services (NERC, NESE, OSN) for domain researchers and education that are operated and facilitated by university research IT, a national testbed for cloud research (OCT), projects to enable scientific and medical researchers (Biogrids, ChRiS), and projects to engage the open source (OI labs) and system research (Red Hat Collaboratory, i-Scale) communities. This talk will describe the status of the open production cloud and discuss how it has became a laboratory for cloud research and innovation, resulting in 100s of publications, contributions to open source software, and collaborations between researchers, open source developers, and production operations staff.

9:30 AM - 10:00 AM

Coffee Break

10:00 AM - 12:00 PM

Session 3: Experiments
Session chair: Odej Kao

  • CloudFactory: An Open Toolkit to Generate Production-like Workloads for Cloud Infrastructures (Research Track)
    Pierre Jacquet, Thomas Ledoux, and Romain Rouvoy
  • Towards a Benchmark for Fog Data Processing (Research Track)
    Tobias Pfandzelter and David Bermbach
  • Studying the Energy Consumption of Stream Processing Engines in the Cloud (Research Track)
    Govind K.P., Guillaume Pierre, and Romain Rouvoy
  • Inducing Huge Tail Latency on a MongoDB deployment (Research Track)
    Remo Andreoli and Tommaso Cucinotta

12:00 PM - 1:30 PM

Lunch

1:30 PM - 3:00 PM

Session 4: Machine Learning
Session chair: Wes Lloyd

  • Learning to Score: Tuning Cluster Schedulers through Reinforcement Learning (Research Track)
    Martin Asenov, Qiwen Deng, Gingfung Yeung, and Adam Barker
  • A Reinforcement Learning Approach for Performance-aware Reduction in Power Consumption of Data Center Compute Nodes (Research Track)
    Akhilesh Raj, Swann Perarnau, and Aniruddha Gokhale
  • Voda: A GPU Scheduling Platform for Elastic Deep Learning in Kubernetes Cluster (Research Track)
    Tsung-Tso Hsieh and Che-Rung Lee
  • Exploring the Impact of Serverless Computing on Peer To Peer Training Machine Learning (Research Track)
    Amine Barrak, Ranim Trabelsi, Fehmi Jaafar, and Fabio Petrillo

3:00 PM - 3:30 PM

Coffee break

3:30 PM - 5:00 PM

Session 5: Scheduling
Session chair: Tommaso Cucinotta

  • Cost-Effective Scheduling for Kubernetes in the Edge-to-Cloud Continuum (Research Track)
    Samuel Rac and Mats Brorsson
  • Vela: A 3-Phase Distributed Scheduler for the Edge-Cloud Continuum (Research Track)
    Thomas Werner Pusztai, Stefan Nastic, Philipp Raith, Schahram Dustdar, Deepak Vij, and Ying Xiong
  • Multi-Objective Workflow Scheduling to Serverless Architecture in a Multi-Cloud Environment (Industry Track)
    Manju Ramesh, Dheeraj Chahal, Chetan Phalak, and Rekha Singhal

9:30 AM - 11:30 AM

Session 6: Emergent Opportunities
Session chair: Stefan Schulte

  • REFORM: Increase alerts value using data driven approach (Industry Track)
    Anupama Jagannathan, Chris Dye, Karthick Rajamani, Chris Galtenberg, Benjamin Luong, and Egan Ford
  • Service Function Chaining Implementation using VNFs and CNFs (Industry Track)
    Abdullah Bittar, Ziqiang Wang, and Changcheng Huang
  • Evaluation of Data Enrichment Methods for Distributed Stream Processing Systems (Research Track)
    Dominik Scheinert, Fabian Casares, Morgan Geldenhuys, Kevin Styp-Rekowski, and Odej Kao
  • A Crowdsensing Approach for Deriving Surface Quality of Cycling Infrastructure (Research Track)
    Ahmet-Serdar Karakaya, Leonard Thomas, Denis Koljada, and David Bermbach

12:00 PM - 1:00 PM

Lunch